Merge branch 'master' of git.metadada.xyz:machinelistening/curriculum

PUBLISH.trigger.md

@@ -1,13 +1,13 @@
-1 23456 7 8 9 10
-a
-jp 3 1
-zoe publishing new interview . 
+don't gaslight me bro
+a2`121321111
+jp 3 1d2131
+zoe publishing new interview . k11
 JP links to faculty profiles
-zd adding socials, bio for bridget, spelling for xiaochang
+zd adding socials, bio for bridget, spe1lling for xiaochang
 fixing ugly links for socials
-socials again
-interview url
-la bios for singapore
+socials again2
+interview urljjn
+la bios for singaporeg
 singapore contributors to main contributors list and ntu link at top of "next episode"
 as above with edits
 12

content/curriculum/index.md

@@ -5,6 +5,7 @@ has_topics:
     "against-the-coming-world-of-listening-machines.md",
     "lessons-in-how-not-to-be-heard.md",
     "listening-with-the-pandemic.md",
+    "improvisation-and-control.md",
     "interviews.md",
   ]
 description: "Machine Listening Curriculum: A platform for collective listening, thought, and artistic production: a critical counterpoint to all the solutionists, VCs, militarists and industry boosters intent on 'empowering machines with the sense of hearing'."
@@ -38,7 +39,7 @@ MACHINE LISTENING, A CURRICULUM is an evolving resource, comprising existing and
 
 Amidst oppressive and extractive forms of state and corporate listening, practices of collaborative study, experimentation and resistance will, we hope, enable us to develop strategies for recalibrating our relationships to machine listening, whether through technological interventions, alternative infrastructures, new behaviors, or political demands. With so many cultural producers – whose work and research is crucial for this kind of project – thrown into deeper precarity and an uncertain future by the unfolding pandemic, we also hope that this curriculum will operate as a quasi-institution: a site of collective learning about and mobilisation against the coming world of listening machines.
 
-A curriculum is also a technology, a tool for supporting and activating learning. And this one is open source. It has been built on a platform developed by [Pirate Care](https://syllabus.pirate.care/) for their own experiments in open pedagogy. We encourage everyone to freely use it to learn and organise processes of learning and to freely adapt, rewrite and expand it to reflect their own experience and serve their own pedagogies. As the curriculum unfolds, these resources will expand: {{< nosup black >}}[event documentation]( https://machinelistening.exposed/_new/_preview/curriculum/#schedule "Video from past curriculum events."){{< /nosup >}}, {{< nosup black >}}[interviews]( https://machinelistening.exposed/_new/_preview/topic/interviews/ "Recorded interviews with contributors."){{< /nosup >}}, and {{< nosup black >}}[library]( https://machinelistening.exposed/_new/_preview/library/BROWSE_LIBRARY.html "A growing collection of Machine Listening related texts and recordings."){{< /nosup >}}.
+A curriculum is also a technology, a tool for supporting and activating learning. And this one is open source. It has been built on a platform developed by [Pirate Care](https://syllabus.pirate.care/) for their own experiments in open pedagogy. We encourage everyone to freely use it to learn and organise processes of learning and to freely adapt, rewrite and expand it to reflect their own experience and serve their own pedagogies. As the curriculum unfolds, these resources will expand: {{< nosup black >}}[event documentation](https://www.youtube.com/playlist?list=PLrN3t2eBJmgtmKg2Gzsl_V2tvDkoOAhU1 "Video from past curriculum events."){{< /nosup >}}, {{< nosup black >}}[interviews]( https://machinelistening.exposed/_new/_preview/topic/interviews/ "Recorded interviews with contributors."){{< /nosup >}}, and {{< nosup black >}}[library]( https://machinelistening.exposed/_new/_preview/library/BROWSE_LIBRARY.html "A growing collection of Machine Listening related texts and recordings."){{< /nosup >}}.
 
 ![ML](static/images/03.gif)
 
@@ -58,12 +59,12 @@ Across three days at the start of October 2020, we came together to investigate
 {{< nosup >}}[(documentation)](https://www.youtube.com/watch?v=aM5SVoMBnKI){{< /nosup >}}
 
 3. Sun, 04. October 2020: ![](topic:listening-with-the-pandemic.md)
-{{< nosup >}}[( documentation)](https://www.youtube.com/watch?v=vuNmI9Xdgpo){{< /nosup >}}
+{{< nosup >}}[(documentation)](https://www.youtube.com/watch?v=vuNmI9Xdgpo){{< /nosup >}}
 
 **Liquid Architecture x NTU CCA Singapore**
 Sat, 13. March 2021
 
-4. Sat, 13. March 2021: Improvisation and Control {{< nosup black >}}[(zoom)](https://us02web.zoom.us/webinar/register/WN_AYsgDHKhRNGa517hPb3gYA){{< /nosup >}}
+4. Sat, 13. March 2021: Improvisation and Control
 
 As part of _Free Jazz III_, [Improvisation and Control](http://ntu.ccasingapore.org/events/liquid-architecturemachine-listening-a-curriculum/) explores machine listening’s history in computer music, and the evolving dynamics between improvisation and control.
 

content/experiment/us-listening-to-machines-listening-to-us.md

@@ -6,10 +6,9 @@ title: "Us Listening to Machines Listening to Us"
 
 A score and experiment for [Zoom](https://zoom.us/), by [Mattin](http://mattin.org/)
 
-Commissioned for and first performed at Machine Listening: Improvisation and Control, 13 March 2021
-{{< nosup >}}[Documentation](https://youtu.be/EZvK8atIlnA?t=4033){{< /nosup >}}
+Commissioned for and first performed at [Machine Listening: Improvisation and Control](https://www.youtube.com/watch?v=EZvK8atIlnA&list=PLrN3t2eBJmgtmKg2Gzsl_V2tvDkoOAhU1&index=4&t=6180s), 13 March 2021
 
-Using a simple instructional score, I propose for my presentation to do a collective improvisation involving the audience using the audio settings of ZOOM. The idea will be to play with the algorithms that regulate AEC (Acoustic Echo Cancellation), AGC (Automatic Gain Control) and NS (Noise Suppression). By doing this we hopefully get more traction on the specific forms of mediation that involves sound within the reality of Zoom so prevalent these days.
+"Using a simple instructional score, I propose for my presentation to do a collective improvisation involving the audience using the audio settings of ZOOM. The idea will be to play with the algorithms that regulate AEC (Acoustic Echo Cancellation), AGC (Automatic Gain Control) and NS (Noise Suppression). By doing this we hopefully get more traction on the specific forms of mediation that involves sound within the reality of Zoom so prevalent these days." Mattin
 
 
 ## Score

content/interview/alex-ahmed.md

@@ -0,0 +1,23 @@
+---
+title: "Alex Ahmed"
+Description: "[Alex](https://scholar.google.com/citations?user=Gc8T8LkAAAAJ&hl=en) talks to us about [Project Spectra](https://github.com/project-spectra), an online, community-based, free and open source software application for transgender voice training. We discuss speech pathology and the politics of pitch, along with the importance of grass-roots led tech projects and community-centred design."
+aliases: []
+author: "Machine Listening"
+date: "2020-09-21T00:00:00-05:00"
+episode_image: "images/ml.gif"
+explicit: "no"
+hosts: ["james-parker", "joel-stern", "sean-dockray"]
+images: ["images/ml.gif"]
+news_keywords: []
+podcast_duration: "50:26"
+podcast_file: "https://machinelistening.exposed/library/Alex%20Ahmed/Alex%20Ahmed%20(22)/Alex%20Ahmed%20-%20Alex%20Ahmed.mp3"
+podcast_bytes: ""
+youtube: ""
+categories: []
+series: []
+tags: []
+transcript: "http://git.metadada.xyz/machinelistening/curriculum/src/branch/master/content/transcript/ahmed.md"
+---
+
+
+

content/interview/andre-dao.md

@@ -0,0 +1,20 @@
+---
+title: "André Dao"
+Description: "[André](https://andredao.com/) talks to us about [UN Global Pulse](https://www.unglobalpulse.org/), the UN's big data initiative, and in particular one program which 'uses machine Learning to analyse radio content in Uganda'. We discuss the increasing entanglements of big tech, the UN and human rights discourse more broadly, as well as an emergent _right to be counted_."
+aliases: []
+author: "Machine Listening"
+date: "2020-09-04T00:00:00-05:00"
+episode_image: "images/ml.gif"
+explicit: "no"
+hosts: ["james-parker", "joel-stern", "sean-dockray"]
+images: ["images/ml.gif"]
+news_keywords: []
+podcast_duration: "62:34"
+podcast_file: "https://machinelistening.exposed/library/Andre%20Dao/Andre%20Dao%20(26)/Andre%20Dao%20-%20Andre%20Dao.mp3" 
+podcast_bytes: ""
+youtube: ""
+categories: []
+series: []
+tags: []
+transcript: "http://git.metadada.xyz/machinelistening/curriculum/src/branch/master/content/transcript/dao.md"
+---

content/interview/angie-abdilla.md

@@ -0,0 +1,23 @@
+---
+title: "Angie Abdilla"
+Description: "Angie talks to us about [Old Ways, New](https://oldwaysnew.com/), the Indigenous owned and led social enterprise she founded, based on Gadigal land in Redfern, Sydney. We discuss [Decolonising the Digital](http://ojs.decolonising.digital/index.php/decolonising_digital), Country Centered Design, a methodology which applies Indigenous design principles to the development of technologies for places, spaces and experiences, and how this contrasts with the ‘placelessness’ on which so many machine learning/listening systems are based."
+aliases: []
+author: "Machine Listening"
+date: "2020-09-01T00:00:00-05:00"
+episode_image: "images/ml.gif"
+explicit: "no"
+hosts: ["james-parker", "joel-stern", "sean-dockray"]
+images: ["images/ml.gif"]
+news_keywords: []
+podcast_duration: "38:49"
+podcast_file: "https://machinelistening.exposed/library/Angie%20Abdilla/Angie%20Abdilla%20(31)/Angie%20Abdilla%20-%20Angie%20Abdilla.mp3"
+podcast_bytes: ""
+youtube: ""
+categories: []
+series: []
+tags: []
+transcript: "http://git.metadada.xyz/machinelistening/curriculum/src/branch/master/content/transcript/abdilla.md"
+---
+
+
+

content/interview/feldman-li-mills-pfeiffer.md

@@ -4,7 +4,7 @@ Description: "[Mara](http://maramills.org/), [Xiaochang](https://comm.stanford.e
 Description2: "Mara, Xiaochang, Jessica and Michelle talk us through the history and politics of machine listening, from 'affect recognition' and the 'statistical turn' in ASR to automated accent detection at the German border, voiceprints and the 'assistive pretext'. This is an expansive conversation with an amazing group of scholars, who share a common connection to the Media, Culture, and Communications department at NYU, founded by Neil Postman in 1971 at the urging of Marshall McLuhan."
 aliases: []
 author: "Machine Listening"
-date: "2021-03-01T00:00:00-05:00"
+date: "2020-02-19T00:00:00-05:00"
 episode_image: "images/ml.gif"
 explicit: "no"
 hosts: ["james-parker", "joel-stern", "sean-dockray"]

content/interview/lauren-lee-mccarthy.md

@@ -0,0 +1,21 @@
+---
+title: "Lauren Lee McCarthy"
+Description: "[Lauren](https://lauren-mccarthy.com/) talks us through some of her many works concerned with smart speakers, machine listening and social relationships in the midst of surveillance, automation, and algorithmic living. We discuss: [LAUREN](https://lauren-mccarthy.com/LAUREN), for which she attempted to become a human version of Alexa, [SOMEONE](https://lauren-mccarthy.com/SOMEONE), which won her the [Prix Ars Electronica 2020 / Interactive Art +](https://ars.electronica.art/homedelivery/en/winners-prix-interactive-art/), and a range of related works and political questions."
+aliases: []
+author: "Machine Listening"
+date: "2020-09-22T00:00:00-05:00"
+episode_image: "images/ml.gif"
+explicit: "no"
+hosts: ["james-parker", "joel-stern", "sean-dockray"]
+images: ["images/ml.gif"]
+news_keywords: []
+podcast_duration: "38:55"
+podcast_file: "https://machinelistening.exposed/library/Lauren%20Lee%20McCarthy/Lauren%20Lee%20McCarthy%20(23)/Lauren%20Lee%20McCarthy%20-%20Lauren%20Lee%20McCarthy.mp3"
+podcast_bytes: ""
+youtube: ""
+categories: []
+series: []
+tags: []
+transcript: "http://git.metadada.xyz/machinelistening/curriculum/src/branch/master/content/transcript/mccarthy.md"
+---
+

content/interview/li.md

@@ -0,0 +1,20 @@
+---
+title: "Jùnchéng Billy Lì"
+Description: "[Billy](https://lijuncheng16.github.io/index.html) tells us about his research on 'adversarial music', and in particular an attempt to produce a ['Real World Audio Adversary Against Wake-word Detection Systems'](https://machinelistening.exposed/library/BROWSE_LIBRARY.html#/book/fac6c1a2-946f-43c4-83f5-e54fd7185c18) for Amazon Alexa."
+aliases: []
+author: "Machine Listening"
+date: "2020-09-11T00:00:00-05:00"
+episode_image: "images/ml.gif"
+explicit: "no"
+hosts: ["james-parker", "joel-stern", "sean-dockray"]
+images: ["images/ml.gif"]
+news_keywords: []
+podcast_duration: "37:32"
+podcast_file: "https://machinelistening.exposed/library/Juncheng%20Billy%20Li/Juncheng%20Billy%20Li%20(25)/Juncheng%20Billy%20Li%20-%20Juncheng%20Billy%20Li.mp3"
+podcast_bytes: ""
+youtube: ""
+categories: []
+series: []
+tags: []
+transcript: "http://git.metadada.xyz/machinelistening/curriculum/src/branch/master/content/transcript/billyli.md"
+---

content/interview/parker-guffond.md

@@ -0,0 +1,20 @@
+---
+title: "James Parker (w Jasmine Guffond)"
+Description: "This is the first of three radio shows as part of [Jasmine's](https://jasmineguffond.bandcamp.com/album/microphone-permission) guest residency at Noods Radio. It features an interview with [James](https://law.unimelb.edu.au/about/staff/james-parker) about his research on machine listening, this curriculum, the project with Unsound, and a selection of electronic music."
+aliases: []
+author: "Machine Listening"
+date: "2020-09-01T00:00:00-05:00"
+episode_image: "images/ml.gif"
+explicit: "no"
+hosts: ["james-parker", "joel-stern", "sean-dockray"]
+images: ["images/ml.gif"]
+news_keywords: []
+podcast_duration: "59:57"
+podcast_file: "https://machinelistening.exposed/library/James%20Parker/James%20Parker%20(30)/James%20Parker%20-%20James%20Parker.mp3"
+podcast_bytes: ""
+youtube: ""
+categories: []
+series: []
+tags: []
+transcript: "http://git.metadada.xyz/machinelistening/curriculum/src/branch/master/content/transcript/parker.md"
+---

content/interview/stefan-maier.md

@@ -0,0 +1,26 @@
+---
+title: "Stefan Maier"
+Description: "[Stefan's](http://stefanmaier.studio/info/) 2018 [dossier on machine listening for Technosphere](https://technosphere-magazine.hkw.de/p/1-WaveNet-On-Machine-and-Machinic-Listening-a2mD8xYCxtsLqoaAnTGUbn) puts the work of artists like George Lewis, Jennifer Walshe, Florian Hecker, and Maryanne Amacher into conversation with Google's wavenet. We talk about these and other works along with Stefan's own compositions which treat machine listening as a prepared instrument, ready to be detourned."
+aliases: []
+author: "Machine Listening"
+date: "2020-09-11T00:00:00-05:00"
+episode_image: "images/ml.gif"
+explicit: "no"
+hosts: ["james-parker", "joel-stern", "sean-dockray"]
+images: ["images/ml.gif"]
+news_keywords: []
+podcast_duration: "38:55"
+podcast_file: "https://machinelistening.exposed/library/Stefan%20Maier/Stefan%20Maier%20(21)/Stefan%20Maier%20-%20Stefan%20Maier.mp3"
+podcast_bytes: ""
+youtube: ""
+categories: []
+series: []
+tags: []
+transcript: "http://git.metadada.xyz/machinelistening/curriculum/src/branch/master/content/transcript/maier.md"
+---
+
+
+
+
+
+11 September

content/interview/strengers-kennedy.md

@@ -0,0 +1,20 @@
+---
+title: "Yolande Strengers and Jenny Kennedy"
+Description: "[Yolande](https://research.monash.edu/en/persons/yolande-strengers) and [Jenny](https://www.jennykennedy.net/) provide a “reboot” manifesta in their book [_The Smart Wife: Why Siri, Alexa, and Other Smart Home Devices Need a Feminist Reboot_](https://www.audible.com.au/pd/The-Smart-Wife-Audiobook/1705255280), which lays out their proposals for improving the design and social effects of digital voice assistants, social robots, sex robots, and other AI arriving in the home."
+aliases: []
+author: "Machine Listening"
+date: "2020-10-11T00:00:00-05:00"
+episode_image: "images/ml.gif"
+explicit: "no"
+hosts: ["james-parker", "joel-stern", "sean-dockray"]
+images: ["images/ml.gif"]
+news_keywords: []
+podcast_duration: "38:55"
+podcast_file: "https://machinelistening.exposed/library/Yolande%20Strengers/Yolande%20Strengers%20and%20Jenny%20Kennedy%20(10)/Yolande%20Strengers%20and%20Jenny%20Ken%20-%20Yolande%20Strengers.mp3"
+podcast_bytes: ""
+youtube: ""
+categories: []
+series: []
+tags: []
+transcript: "http://git.metadada.xyz/machinelistening/curriculum/src/branch/master/content/transcript/smartwife.md"
+---

content/interview/vladan-joler.md

@@ -0,0 +1,20 @@
+---
+title: "Vladan Joler"
+Description: "Vladan walks us through [Anatomy of an AI System](https://anatomyof.ai/), his 2018 work with [Kate Crawford](https://www.katecrawford.net/), which diagrams the Amazon Echo as an anatomical map of human labor, data and planetary resources. We talk about the politics of visibility and method as well as Vladan's work with [Share Lab](https://labs.rs/en/), 'where indie data punk meets media theory pop to investigate digital rights blues'."
+aliases: []
+author: "Machine Listening"
+date: "2020-09-01T00:00:00-05:00"
+episode_image: "images/ml.gif"
+explicit: "no"
+hosts: ["james-parker", "joel-stern", "sean-dockray"]
+images: ["images/ml.gif"]
+news_keywords: []
+podcast_duration: "55:53"
+podcast_file: "https://machinelistening.exposed/library/Vladan%20Joler/Vladan%20Joler%20(part%202)%20(29)/Vladan%20Joler%20(part%202)%20-%20Vladan%20Joler.mp3"
+podcast_bytes: ""
+youtube: ""
+categories: []
+series: []
+tags: []
+transcript: "http://git.metadada.xyz/machinelistening/curriculum/src/branch/master/content/transcript/joler.md"
+---

content/session/ccantu-2021.md

@@ -0,0 +1,21 @@
+---
+title: "13 March 2021"
+---
+
+Staged as part of the [NTU CCA Singapore’s](https://ntu.ccasingapore.org/) recurring Free Jazz program, this episode of Machine Listening focused on the complex and evolving dialectic between improvisation and control, via a detour into the experimental computer music laboratories of the 1980s and 90s where the term ‘machine listening’ first begins to circulate.
+
+Runsheet:
+
+* {{<nosup>}}[Machine Listening, Essay I: Interactive (music) systems](https://www.youtube.com/watch?v=EZvK8atIlnA&t=120s){{</nosup>}}
+* {{<nosup>}}[Jessica Feldman](https://www.youtube.com/watch?v=EZvK8atIlnA&t=932s){{</nosup>}}
+* {{<nosup>}}[Luca Lum](https://www.youtube.com/watch?v=EZvK8atIlnA&t=2520s){{</nosup>}}
+* {{<nosup>}}[Machine Listening, Essay II: Rainbow Family](https://www.youtube.com/watch?v=EZvK8atIlnA&t=3242s) {{</nosup>}}
+* {{<nosup>}}[Mattin](https://www.youtube.com/watch?v=EZvK8atIlnA&t=4039s) {{</nosup>}}
+* {{<nosup>}}[Bani Haykal & Lee Weng Choy](https://www.youtube.com/watch?v=EZvK8atIlnA&t=5152s) {{</nosup>}}
+* {{<nosup>}}[Machine Listening, Essay III: DARPA Improv](https://www.youtube.com/watch?v=EZvK8atIlnA&t=6196s) {{</nosup>}}
+* {{<nosup>}}[Bridget Chappell](https://www.youtube.com/watch?v=EZvK8atIlnA&t=6802s) {{</nosup>}}
+* {{<nosup>}}[Lee Gamble](https://www.youtube.com/watch?v=EZvK8atIlnA&t=7700s) {{</nosup>}}
+
+Full documentation:
+
+{{< youtube EZvK8atIlnA >}}

content/session/unsound-2020-session-1.md

@@ -1,18 +1,21 @@
 ---
-title: "Unsound 02.10.2020"
+title: "02 Oct 2020"
 ---
-
-**Fri, 02. October 2020**
+Across three days at the start of October 2020, we teamed up with [Unsound](https://www.unsound.pl/) to investigate the implications of the coming world of listening machines in both its dystopian and utopian dimensions. Comprising a montage of presentations, performance, sound, video, music and experiments in listening featuring contributors from around the world, the online gatherings were divided into three sections, open to all. This was session 1.
 
 Runsheet:
 
-* {{<nosup>}}[Sean Dockray, _Always Learning, Everywhere (2020)_](http://www.youtube.com/watch?v=ddnvR7IGpes&t=1m10s) {{</nosup>}}
-* {{<nosup>}}[Kate Crawford](http://www.youtube.com/watch?v=aM5SVoMBnKI%3Ft%3D42m47s&t=42m47s){{</nosup>}}
-* {{<nosup>}}[Stefan Maier](http://www.youtube.com/watch?v=ddnvR7IGpes&t=83m28s) {{</nosup>}}
-* {{<nosup>}}[Jennifer Walshe](http://www.youtube.com/watch?v=ddnvR7IGpes&t=142m1s) {{</nosup>}}
-* {{<nosup>}}[Tom Smith](https://liquidarchitecture.org.au/artists/tom-smith) {{</nosup>}}
+* {{<nosup>}}[Sean Dockray, _Always Learning, Everywhere (2020)_](https://www.youtube.com/watch?v=iUbglqQLdrI&t=74s) {{</nosup>}}
+* {{<nosup>}}[Kate Crawford](https://www.youtube.com/watch?v=iUbglqQLdrI&t=1570s){{</nosup>}}
+* {{<nosup>}}[Sean Dockray, _Learning from YouTube (2018)_](https://www.youtube.com/watch?v=iUbglqQLdrI&t=2892s) {{</nosup>}}
+* {{<nosup>}}[Lauren Lee McCarthy, _I heard TALKING IS DANGEROUS (2020)_](https://www.youtube.com/watch?v=iUbglqQLdrI&t=3792s) {{</nosup>}}
+* {{<nosup>}}[Stefan Maier](https://www.youtube.com/watch?v=iUbglqQLdrI&t=5008s) {{</nosup>}}
+* {{<nosup>}}[Hito Steyerl, _The City of Broken Windows (2018)_](https://www.youtube.com/watch?v=iUbglqQLdrI&t=6460s) {{</nosup>}}
+* {{<nosup>}}[André Dao, _No Voice Left Behind (2020)_](https://www.youtube.com/watch?v=iUbglqQLdrI&t=7515s) {{</nosup>}}
+* {{<nosup>}}[Jennifer Walshe](https://www.youtube.com/watch?v=iUbglqQLdrI&t=8521s) {{</nosup>}}
+* {{<nosup>}}[Tom Smith](https://machinelistening.exposed/experiment/top-ten/) {{</nosup>}}
 
 
-[full documentation](https://www.youtube.com/watch?v=ddnvR7IGpes "Unsound Machine Listening Session I")
+Full documentation:
 
-{{< youtube ddnvR7IGpes >}}
+{{< youtube iUbglqQLdrI >}}

content/session/unsound-2020-session-2.md

@@ -1,19 +1,19 @@
 ---
-title: "Unsound Machine Listening Session II"
+title: "03 Oct 2020"
 ---
-# Lessons in how (not) to be heard
-
-Sat, 03. October 2020
+Across three days at the start of October 2020, we teamed up with [Unsound](https://www.unsound.pl/) to investigate the implications of the coming world of listening machines in both its dystopian and utopian dimensions. Comprising a montage of presentations, performance, sound, video, music and experiments in listening featuring contributors from around the world, the online gatherings were divided into three sections, open to all. This was session 2.
 
 Runsheet:
 
-* {{<nosup>}}[Halcyon Lawrence](http://www.youtube.com/watch?v=aM5SVoMBnKI%3Ft%3D42m47s&t=26m45s){{</nosup>}}
-* {{<nosup>}}[Joel Spring and Jazz Money](http://www.youtube.com/watch?v=aM5SVoMBnKI%3Ft%3D42m47s&t=42m47s ){{</nosup>}}
-* {{<nosup>}}[Karolina Iwańska, Panoptykon Foundation](http://www.youtube.com/watch?v=aM5SVoMBnKI&t=66m18s) {{</nosup>}}
-* {{<nosup>}}[Alex A. Ahmed](http://www.youtube.com/watch?v=aM5SVoMBnKI&t=97m51s) {{</nosup>}}
-* {{<nosup>}}[Mat Dryhurst](http://www.youtube.com/watch?v=aM5SVoMBnKI&t=119m51s) {{</nosup>}}
+* {{<nosup>}}[Ad-versarial Audio](https://youtu.be/aS2Fp3W8l6A?t=1){{</nosup>}}
+* {{<nosup>}}[Halcyon Lawrence](https://www.youtube.com/watch?v=aS2Fp3W8l6A&t=1605s){{</nosup>}}
+* {{<nosup>}}[Joel Spring and Jazz Money](https://www.youtube.com/watch?v=aS2Fp3W8l6A&t=2567s){{</nosup>}}
+* {{<nosup>}}[Karolina Iwańska, Panoptykon Foundation](https://www.youtube.com/watch?v=aS2Fp3W8l6A&t=3978s) {{</nosup>}}
+* {{<nosup>}}[Lawrence Abu Hamdan](https://www.youtube.com/watch?v=aS2Fp3W8l6A&t=4628s) {{</nosup>}}
+* {{<nosup>}}[Alex A. Ahmed](https://www.youtube.com/watch?v=aS2Fp3W8l6A&t=5871s) {{</nosup>}}
+* {{<nosup>}}[Mat Dryhurst](https://www.youtube.com/watch?v=aS2Fp3W8l6A&t=7191s) {{</nosup>}}
 
 
-[full documentation](https://www.youtube.com/watch?v=aS2Fp3W8l6A "Unsound Machine Listening Session II")
+Full documentation:
 
-{{< youtube ddnvR7IGpes >}}
+{{< youtube aS2Fp3W8l6A >}}

content/session/unsound-2020-session-3.md

@@ -1,17 +1,18 @@
 ---
-title: "Unsound Machine Listening Session III"
+title: "04 Oct 2020"
 ---
-# Listening with the pandemic
-
-Sun, 03. October 2020
+Across three days at the start of October 2020, we teamed up with [Unsound](https://www.unsound.pl/) to investigate the implications of the coming world of listening machines in both its dystopian and utopian dimensions. Comprising a montage of presentations, performance, sound, video, music and experiments in listening featuring contributors from around the world, the online gatherings were divided into three sections, open to all. This was session 3.
 
 Runsheet:
-* {{<nosup>}}[Sean Dockray](http://www.youtube.com/watch?v=vuNmI9Xdgpo&t=22m35s){{</nosup>}}
-* {{<nosup>}}[Yeshimabeit Milner](http://www.youtube.com/watch?v=vuNmI9Xdgpo&t=46m5s){{</nosup>}}
-* {{<nosup>}}[Thao Phan](http://www.youtube.com/watch?v=vuNmI9Xdgpo&t=87m21s) {{</nosup>}}
-* {{<nosup>}}[Vladan Joler](http://www.youtube.com/watch?v=vuNmI9Xdgpo&t=87m21s) {{</nosup>}}
-* {{<nosup>}}[Andrew Brooks](http://www.youtube.com/watch?v=vuNmI9Xdgpo&t=87m21s) {{</nosup>}}
-* {{<nosup>}}[Shannon Mattern](http://www.youtube.com/watch?v=vuNmI9Xdgpo&t=154m40s) {{</nosup>}}
+* {{<nosup>}}[Sean Dockray](https://www.youtube.com/watch?v=7mcBE-qTcVI&list=PLrN3t2eBJmgtmKg2Gzsl_V2tvDkoOAhU1&index=3&t=1355s){{</nosup>}}
+* {{<nosup>}}[Yeshimabeit Milner, Data For Black Lives](https://www.youtube.com/watch?v=7mcBE-qTcVI&list=PLrN3t2eBJmgtmKg2Gzsl_V2tvDkoOAhU1&index=3&t=2769s){{</nosup>}}
+* {{<nosup>}}[Mark Andrejevic](https://www.youtube.com/watch?v=7mcBE-qTcVI&list=PLrN3t2eBJmgtmKg2Gzsl_V2tvDkoOAhU1&index=3&t=4040s){{</nosup>}}
+* {{<nosup>}}[Thao Phan](https://www.youtube.com/watch?v=7mcBE-qTcVI&list=PLrN3t2eBJmgtmKg2Gzsl_V2tvDkoOAhU1&index=3&t=5241s) {{</nosup>}}
+* {{<nosup>}}[Vladan Joler](https://www.youtube.com/watch?v=7mcBE-qTcVI&list=PLrN3t2eBJmgtmKg2Gzsl_V2tvDkoOAhU1&index=3&t=6142s) {{</nosup>}}
+* {{<nosup>}}[Andrew Brooks](https://www.youtube.com/watch?v=7mcBE-qTcVI&list=PLrN3t2eBJmgtmKg2Gzsl_V2tvDkoOAhU1&index=3&t=7434s) {{</nosup>}}
+* {{<nosup>}}[Shannon Mattern](https://www.youtube.com/watch?v=7mcBE-qTcVI&list=PLrN3t2eBJmgtmKg2Gzsl_V2tvDkoOAhU1&index=3&t=9280s) {{</nosup>}}
+
 
+Full documentation:
 
-[full documentation](https://www.youtube.com/watch?v=7mcBE-qTcVI "Unsound Machine Listening Session III")
+{{< youtube 7mcBE-qTcVI >}}

content/topic/against-the-coming-world-of-listening-machines.md

@@ -128,25 +128,25 @@ Another response would be to say that when or if machines listen, they listen ![
 
 [^Cella, Serizel, Ellis]: ![](bib:7c769ce6-5e9e-40d3-96ef-1838a7f57365)
 [^alper]: Meryl Alper, _Giving Voice: Mobile Communication, Disability_, and Inequality (MIT Press, 2017)
-[^rowe]: LIBRARY Robert Rowe, _Interactive music systems: Machine listening and composing._ Cambridge,
+[^rowe]: Robert Rowe, [_Interactive music systems: Machine listening and composing._](https://wp.nyu.edu/robert_rowe/text/interactive-music-systems-1993/) Cambridge,
 MA: The MIT Press (1993)
 [^maier_audio_1]: Stefan Maier, [_Machine Listening_](https://technosphere-magazine.hkw.de/p/Machine-Listening-kmgQVZVaQeugBaizQjmZnY), Technosphere Magazine (2018); Interview with [Stefan Maier](http://stefanmaier.studio/info/) on September 11, 2020
-[^li and mills]: LIBRARY Xiaochang Li and Mara Mills, "Vocal Features: From Identification to Speech Recognition by Machine" 60(2) _Technology and Culture_ (2019) pp.129-S160 DOI: https://doi.org/10.1353/tech.2019.0066
+[^li and mills]: Xiaochang Li and Mara Mills, "Vocal Features: From Identification to Speech Recognition by Machine" 60(2) _Technology and Culture_ (2019) pp.129-S160 DOI: https://doi.org/10.1353/tech.2019.0066
 [^intelligent_audio_analysis]: ![](bib:827d1f44-5a35-4278-a527-4df67e5ba321)
-[^virtanen]: LIBRARY Virtanen et al, _Computational Analysis of Sound Scenes and Events_ (Springer, 2017)
-[^lei_mak]: LIBRARY Lei and Mak, "Robust scream sound detection via sound event partitioning"
-[^Abu Hamdan]: Interview with Lawrence Abu Hamdan
+[^virtanen]: ![](bib:7cf99c5d-1a28-44d9-958a-8ff5e9cb4441)
+[^lei_mak]: Lei, B., Mak, MW. "Robust scream sound detection via sound event partitioning. Multimed Tools Appl" 75, 6071–6089 (2016). https://doi.org/10.1007/s11042-015-2555-z 
+[^Abu Hamdan]: Interview with Lawrence Abu Hamdan, publication forthcoming 2021
 [^airplanes]: ![](bib:6676af8a-7a4d-4aa8-af96-f26452f58753)
 [^kathy_audio_1]: Interview with [Kathy Reid](https://blog.kathyreid.id.au) on August 11, 2020
 [^mattern]: Interview with [Shannon Mattern](https://wordsinspace.net/shannon/) on August 18, 2020
 [^andre_audio_1]: Interview with [André Dao](https://andredao.com/) on September 4, 2020
-[^seaver]: LIBRARY Nick Seaver "Captivating algorithms: Recommender systems as traps" 24(4) _Journal of Material Culture_ (2018), 421-436
-[^exemplary projects]: See for instance [Data 4 Black Lives](https://d4bl.org/programs.html), [Feminist Data Manifest-No](https://www.manifestno.com/) [add]
+[^seaver]: ![](bib:d2b1e24c-c800-42b9-ba67-105b0b25efc9)
+[^exemplary projects]: See for instance [Data 4 Black Lives](https://d4bl.org/programs.html), [Feminist Data Manifest-No](https://www.manifestno.com/)
 [^Crawford and Joler]: ![](bib:3f8dd486-3e28-45ef-929f-65086850870e)
-[^goldenfein]: LIBRARY Jake Goldenfein, _Monitoring Laws: Profiling and Identity in the World State_ (Cambridge University Press, 2019) https://doi.org/10.1017/9781108637657
+[^goldenfein]: ![](bib:6e8f7c36-d251-4a07-ac5d-0b938c5f5fee)
 [^mcquillan]: ![](bib:c58be9a5-a599-4a4b-b58f-a07721fc1721)
 [^halcyon_audio_1]: Interview with [Halcyon Lawrence](http://www.halcyonlawrence.com/) on August 31, 2020. See also Thao Phan, "Amazon Echo and the Aesthetics of Whiteneness" 5(1) _Catalyst: Feminism, Theory, Technoscience_ (2019), 1-38.
 [^virilio]: ![](bib:8558647f-101d-43ff-a531-5df8eb87199a) p.53
 [^Faroki, Paglen]: Mark Andrejevic, [Operational Listening (Eavesdropping)](https://youtu.be/OxOKlgsc3_M), recorded on August 10, 2018
-[^Billy Li]: ![](bib:fac6c1a2-946f-43c4-83f5-e54fd7185c18) For a good introduction to adversarialism, see LIBRARY Goodfellow
+[^Billy Li]: ![](bib:fac6c1a2-946f-43c4-83f5-e54fd7185c18) For a good introduction to adversarialism, see ![](bib:bc39dd7f-1dcc-46dc-9f52-6a16b913ff5a)
 [^ys]: ![](bib:26f7b730-9064-464b-b905-fbe63c5d4e4b)

content/topic/improvisation-and-control.md

@@ -1,8 +1,11 @@
 ---
 title: "Improvisation and Control"
-has_experiments: ["us-listening-to-machines-listening-to-us.md"]
+has_experiments: ["us-listening-to-machines-listening-to-us.md",]
+sessions:
+  [
+    "ccantu-2021.md",
+  ]
 ---
-
 # Improvisation and Control
 
 ## Interactive (music) systems
@@ -15,7 +18,7 @@ Is the man playing the computer - like an instrument? Or is he playing with the
 
 The man appears slightly bored, pretending not to be aware of his own performance, exploring the limited freedom offered to him by the machine, which tirelessly repeats the melody again and again, infinitely. We are watching a breakthrough moment in human-computer interaction: the computer is doing what the man wants. But still, the man can only want what the machine can do.
 
-The fantasy of easy, natural interface between man and a computer is captured in a diagram by Andrey Yershov in 1964 titled the 'director agent model of interaction'. The man is meant to be in charge. But [look at the diagram](https://machinelistening.tumblr.com/post/635556111569862656/the-yershov-diagram-1963-as-early-component-of). We can start anywhere we like. Information cycles around and around, in a constant state of transformation from sound to voltage to colored light to wet synapses. All of these possibilities are contained in the schematic figure of the arrow. Which is the director here? And which the agent? The diagram itself cycled between the pages of different publications, including The Architecture Machine, a 1970 book by Nicholas Negroponte [^Negroponte]. Negroponte had set up the Architecture Machine Group at MIT in 1967, which eventually led to his creation of the [MIT Media Lab](https://www.media.mit.edu/).
+The fantasy of easy, natural interface between man and a computer is captured in a diagram by Andrey Yershov in 1964 titled the 'director agent model of interaction'. The man is meant to be in charge. But [look at the diagram](https://machinelistening.tumblr.com/post/635556111569862656/the-yershov-diagram-1963-as-early-component-of). We can start anywhere we like. Information cycles around and around, in a constant state of transformation from sound to voltage to colored light to wet synapses. All of these possibilities are contained in the schematic figure of the arrow. Which is the director here? And which the agent? The diagram itself cycled between the pages of different publications, including The Architecture Machine, a 1970 book by Nicholas Negroponte.[^Negroponte] Negroponte had set up the Architecture Machine Group at MIT in 1967, which eventually led to his creation of the [MIT Media Lab](https://www.media.mit.edu/).
 
 A state of the art, light grey machine is sitting on a white desk. A camera is pointed at it, focused on it. The camera zooms out to reveal a young, white man. Why are we seeing this moment? Why is the camera there to witness it? Judging by the DECstation, the year is probably 1992 or 1993, the location is definitely the MIT Media Lab, and we are looking through a small window into the "demo or die" culture that Negroponte famously instigated there. Demos could excite the general public and impress important visitors. They could attract corporate and government money. The colossal "machine listening" apparatus that we know today has its roots in thousands of demos like this one.
 
@@ -44,22 +47,25 @@ Jessica Feldman's essay, "The Problem of the Adjective," describes a further fro
  
 ## Rainbow Family
 
+{{< yt id="EZvK8atIlnA" yt_start="3614" modestbranding="true" color="white" >}}
+
 A young, black man is sitting at a white desk. In front of him is a bank of Apple II computers, connected to a trio of Yamaha DX-7 synthesisers, and four performers on stage. A woman playing an upright bass starts strumming out a rhythm and the system responds; then the soprano sax, followed by the rest of the ensemble. They are improvising. The machine is listening. The machine is improvising. They are listening.
 
-This is the 1984 premiere of George Lewis' Rainbow Family at IRCAM, Paris, then - as now - a global center for avant-garde composition, computer music and electro-acoustic research. Robert Rowe spent time there in the 1980s, and Lewis would perform at the debut of Rowe's interactive music system, Cypher, at MIT in 1988. The concert was called Hyperinstruments.
+This is the 1984 premiere of George Lewis' [Rainbow Family](https://www.youtube.com/watch?v=i4bS-0tsVEg&feature=emb_title) at [IRCAM](https://www.ircam.fr/), Paris, then - as now - a global center for avant-garde composition, computer music and electro-acoustic research. Robert Rowe spent time there in the 1980s, and Lewis would perform at the debut of Rowe's interactive music system, Cypher, at MIT in 1988. The concert was called Hyperinstruments.[^Rowe]
 
-But this is the first of Lewis' 'interactive virtual orchestra' pieces, in which software designed by Lewis both responds to the sounds of the human performers, and operates independently, according to its own internal processes. There is no 'score'. And Lewis is not, in the language of European 'art music', the piece's 'composer'. Instead, Rainbow Family comprises [quote] 'multiple parallel streams of music generation, emanating from both the computers and the humans—a non-hierarchical, improvisational, subject-subject model of discourse.'
+But this is the first of Lewis' 'interactive virtual orchestra' pieces, in which software designed by Lewis both responds to the sounds of the human performers, and operates independently, according to its own internal processes. There is no 'score'. And Lewis is not, in the language of European 'art music', the piece's 'composer'. Instead, Rainbow Family comprises 'multiple parallel streams of music generation, emanating from both the computers and the humans—a non-hierarchical, improvisational, subject-subject model of discourse.'[^Lewis]
 
-This is not an accident. It is a deliberate aesthetic, technical and political strategy by Lewis, to produce 'a kind of computer music-making embodying African-American aesthetics and musical practices'; a form of human computer collaboration embodying similar ideals to those of the African American musicians' collective AACM - the Association for the Advancement of Creative Musicians. The group was founded in Chicago 1965 and, for Paul Steinbeck, it remains 'the most significant collective organization in the history of jazz and experimental music.'
+This is not an accident. It is a deliberate aesthetic, technical and political strategy by Lewis, to produce 'a kind of computer music-making embodying African-American aesthetics and musical practices';[^Lewis] a form of human computer collaboration embodying similar ideals to those of the African American musicians' collective AACM - the Association for the Advancement of Creative Musicians. The group was founded in Chicago 1965 and, for Paul Steinbeck, it remains 'the most significant collective organization in the history of jazz and experimental music.'[^Steinbeck]
 
-Lewis would later call this AACM-inspired aesthetic 'mulitdominance'. The idea is developed from an essay by the artist and critic Robert L. Douglas. Lewis writes:
+Lewis would later call this AACM-inspired aesthetic 'mulitdominance'. The idea is developed from an essay by the artist and critic Robert L. Douglas.[^Douglas] Lewis writes:
 
->By way of introduction to his theory, Douglas recalls from his art-student days that interviews with “most African-American artists with Eurocentric art training will reveal that they received similar instructions, such as ‘tone down your colors, too many colors’” [11]. Apparently, these “helpful” pedagogical interventions were presented as somehow universal and transcendent, rather than as emanating from a particular culturally or historically situated worldview, or as based in networks of political or social power. Douglas, in observing that [quote] “such culturally narrow aesthetic views would have separated us altogether from our rich African heritage if we had accepted them without question,” goes on to compare this aspect of Eurocentric art training to Eurocentric music training, which in his view does not equip its students to hear music with multidominant rhythmic and melodic elements as anything but “noise,” “frenzy” or perhaps “chaos” [12].
+>By way of introduction to his theory, Douglas recalls from his art-student days that interviews with “most African-American artists with Eurocentric art training will reveal that they received similar instructions, such as ‘tone down your colors, too many colors’”. Apparently, these “helpful” pedagogical interventions were presented as somehow universal and transcendent, rather than as emanating from a particular culturally or historically situated worldview, or as based in networks of political or social power. Douglas, in observing that “such culturally narrow aesthetic views would have separated us altogether from our rich African heritage if we had accepted them without question,” goes on to compare this aspect of Eurocentric art training to Eurocentric music training, which in his view does not equip its students to hear music with multidominant rhythmic and melodic elements as anything but “noise,” “frenzy” or perhaps “chaos”.[^Lewis]
 
 When we listen to Rainbow Family then, Lewis doesn't want us to hear synchronicity, harmony, or even polyphony. He wants us to hear multidominance: as both an aesthetic and a political value, expressed and encapsulated now partly as code. This is a model of human-computer interaction premised on formal equality, difference, independence, and commonality of purpose; a system that is, Lewis explains,
 
->'happy to listen to you and dialog with you, or sometimes ignore you, but the conceptual aspect of it is that it's pretty autonomous. You can't tell it what to do .... So improvisation becomes a negotiation where you have to work with [the system] rather than just be in control.' (Lewis, quoted in Parker 2005, 85)
->'In African American music there is always an instrumentality connected with sounds; you make sounds for pedagogical purposes, to embody history or to tell stories, and so on. (Lewis, quoted in Casserley, 2006)
+>'happy to listen to you and dialog with you, or sometimes ignore you, but the conceptual aspect of it is that it's pretty autonomous. You can't tell it what to do .... So improvisation becomes a negotiation where you have to work with [the system] rather than just be in control.'[^Parker]
+
+>'In African American music there is always an instrumentality connected with sounds; you make sounds for pedagogical purposes, to embody history or to tell stories, and so on.[^Casserley] 
 
 If Rainbow Family is pedagogy then its lesson is surely that computers, and machine listeners in particular, are part of these stories too; that they already were as early as the 1980s. What is being contested in fact is machine listening's soul, the aesthetic and political ideals it both expresses and reproduces, years before the term first began to circulate at and around MIT.
 
@@ -68,26 +74,28 @@ What would an alternate machine listening system modeled along Lewis' lines be l
 
 ## DARPA improv
 
-In this YouTube video, a white woman plays guitar with an Artificial Intelligence. The computer listens and responds. Once again, we find ourselves in a version of Yershov's diagram. It is almost as if she is having a conversation with the AI. She plays, and it listens and responds. She listens to its response and plays some more. This improvised conversation is like the exchange between [video] scientists and aliens in Close Encounters of the Third Kind. It might be playful, or maybe antagonistic - we don't and can't know the meaning even if we could participate in the dialog.
+{{< yt id="EZvK8atIlnA" yt_start="6180" modestbranding="true" color="white" >}}
+
+In [this YouTube video](https://www.youtube.com/watch?v=bRJYcqpbZ9o), a white woman plays guitar with an Artificial Intelligence. The computer listens and responds. Once again, we find ourselves in a version of Yershov's diagram. It is almost as if she is having a conversation with the AI. She plays, and it listens and responds. She listens to its response and plays some more. This improvised conversation is like the exchange between scientists and aliens in Close Encounters of the Third Kind. It might be playful, or maybe antagonistic - we don't and can't know the meaning even if we could participate in the dialog.
 
-The woman playing the guitar also created the alien AI that she improvises with. Still, she can't know the meaning of what it says or what she says in response. The fact that she can't know the meaning and she can't know what response her playing will provoke is precisely what excites her. She wants to create things that she can't quite control. She doesn't play her AI, the way she might play a piano or her guitar, she plays with it.
+[The woman playing the guitar](http://donyaquick.com/) also created the alien AI that she improvises with. Still, she can't know the meaning of what it says or what she says in response. The fact that she can't know the meaning and she can't know what response her playing will provoke is precisely what excites her. She wants to create things that she can't quite control. She doesn't play her AI, the way she might play a piano or her guitar, she plays with it.
 
-This woman and this AI and this amateurish video are part of a research project called MUSICA, short for Musical Interactive Collaborative Agent. The project is funded by DARPA, the US Defense Advanced Research Projects Agency, which also funded some of the early breakthroughs in automatic speech recognition beginning in the 1970s. MUSICA is part of DARPA's 'Communicating with Computers' program. It has two parts: one is called "Composition by Conversation"; the other "Jazz and Musical Improvisation".
+This woman and this AI and this amateurish video are part of a research project called [MUSICA](http://www.musicaresearch.org/), short for Musical Interactive Collaborative Agent. The project is [funded by DARPA](https://www.nbcnews.com/tech/innovation/darpa-wants-make-jazz-playing-robot-can-jam-humans-n449371), the US Defense Advanced Research Projects Agency, which also funded some of the early breakthroughs in automatic speech recognition beginning in the 1970s. MUSICA is part of DARPA's 'Communicating with Computers' program. It has two parts: one is called "Composition by Conversation"; the other "Jazz and Musical Improvisation".
 
 So the question is: why? Why is DARPA into Jazz? Why is it so interested in machines that improvise? What does it think it will learn? What does DARPA imagine improvisation will help it do?
 
 The contemporary battle field is nonlinear. It is often urban. Commanders no longer concentrate their forces along one line or at one point, but disperse them into a 360-degree battlefield. The individual soldier might experience anxiety, even a sense of isolation. They don't simply follow orders, but communicate and flexibly coordinate with other isolated, anxious soldiers. They improvise.
 
-Soldiers improvise with each other, but also with intelligent machines. Another project by the same research group teaches AIs how to infer the internal states of their human teammates, solving problems collaboratively with them, and communicating with them in a socially-aware manner. The practice stage here is Minecraft, where soldiers-in-training can enter a besieged village, [next image] fight a zombie, [image] and are invited to intermittently reflect on their emotional state.
+Soldiers improvise with each other, but also with intelligent machines. [Another project by the same research group](https://ml4ai.github.io/tomcat/) teaches AIs how to infer the internal states of their human teammates, solving problems collaboratively with them, and communicating with them in a socially-aware manner. The practice stage here is Minecraft, where soldiers-in-training can enter a besieged village, fight a zombie, and are invited to intermittently reflect on their emotional state.
 
 Many actors of seemingly different politics have an ideological and tactical investment in improvisation. Improvisation as freedom. Improvisation as multidominance. Improvisation as counter-hegemony; the opposite of control. Mattin has written about "this supposedly self-inherent critical potential of improvisation," and points instead to the way "improvisers embody the precarious qualities of contemporary labor." This is improvisation as corporate agility. Improvisation as zero hours contracts. Improvisation as moving fast and breaking things.
 
 The ability to go off script, off score, to innovate in the moment with what is at hand, becomes a way both of accumulating capital and surviving in the labor market. Evidently for DARPA, improvisation is also the future of combat, so that it is keen to mine the communicative virtuosity of jazz improvisation for whatever secrets it may hold.
 
-In addition to jazz robots, DARPA has two other programs, Improv and Improv2, in which hobbyists (or research labs pretending to be hobbyists) try to create military grade weapons using readily available software and off the shelf technology.
+In addition to jazz robots, DARPA has two other programs, [Improv](https://www.darpa.mil/news-events/2016-03-11) and [Improv2](https://govtribe.com/opportunity/federal-contract-opportunity/improv-2-hr001117s0047), in which hobbyists (or research labs pretending to be hobbyists) try to create military grade weapons using readily available software and off the shelf technology.
 Improvisation is both a technique and a generative source of knowledge to extract. 
 
-But one thing DARPA's Improv program manager says reminds us that their improvisational imaginary has real constraints: "DARPA’s in the surprise business and part of our goal is to prevent surprise." In time, there is no more need for human input in the ensemble. The machine improvises with itself.
+But one thing DARPA's Improv program manager says reminds us that their improvisational imaginary has real constraints: ["DARPA’s in the surprise business and part of our goal is to prevent surprise."](https://spectrum.ieee.org/tech-talk/aerospace/military/darpa-invites-techies-to-turn-offtheshelf-products-into-weapons-in-new-improv-challenge) In time, there is no more need for human input in the ensemble. The machine improvises with itself.
 
 
 # Resources
@@ -98,4 +106,12 @@ But one thing DARPA's Improv program manager says reminds us that their improvis
 [^Ellis]: Archived email exchange between Dan Ellis and [Michael Casey](https://music.dartmouth.edu/people/michael-casey), 28 March 1994. According to Casey, "Dan suggested "Machine Audition", to which I responded that term "audition" was not widely used outside of hearing sciences and medicine, and that it could be a confusing name for a group that was known for working on music--think "music audition". I believe we discussed the word hearing, but I--we?--thought it implied passivity as in "hearing aid", and instead I suggested the name "machine listening" because it had connotations of attention and intelligence, concepts that were of interest to us all at that time. That is what I remember." 
 [^Brand]: ![](bib:f840b2fa-8e2a-48b3-8ad7-1f138313d2b3)
 [^Audioset]: Gemmeke et al. [_Audioset: An Ontology and Human-Labelled Dataset for Audio Events_](https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/45857.pdf)
-[^Feldman]: ![](bib:284b6cc8-1fe8-4d3f-b0b0-d53d4117370b)
+[^Feldman]: ![](bib:284b6cc8-1fe8-4d3f-b0b0-d53d4117370b)
+[^Rowe]: Robert Rowe, [_Interactive Music Systems: Machine Listening and Composing_](https://wp.nyu.edu/robert_rowe/text/interactive-music-systems-1993/) (MIT Press, 1993), 72.
+[^Lewis]: George Lewis, 'Too Many Notes', LEONARDO MUSIC JOURNAL (2000), pp. 33–39, 36-37.
+[^Lewis]: George Lewis, 'Too Many Notes', LEONARDO MUSIC JOURNAL (2000), pp. 33–39, 33.
+[^Steinbeck]: Paul Steinbeck, 'George Lewis' Voyager' _The Routledge Companion to Jazz Studies_ (2018), 261-270, 261.
+[^Lewis]: George Lewis, 'Too Many Notes', LEONARDO MUSIC JOURNAL (2000), pp. 33–39, 34.
+[^Douglas]: Robert L. Douglas, “Formalizing an African-American Aesthetic,” New Art Examiner (June/Summer 1991) pp. 18–24.
+[^Parker]: George Lewis, quoted in Jeff Parker, "George Lewis." BOMB, no. 93 (Fall 2005): 82-88, 85.
+[^Casserley]: George Lewis, quoted in Lawrence Casserley, "Person to ... Person?" _Resonance Magazine_ (1997)

content/topic/interviews.md

@@ -6,67 +6,23 @@ has_experiments: []
 
 {{% interview "feldman-li-mills-pfeiffer.md" %}}
 
+{{% interview "lauren-lee-mccarthy.md" %}}
 
+{{% interview "alex-ahmed.md" %}}
 
-## Lauren Lee McCarthy
+{{% interview "stefan-maier.md" %}}
 
-[Lauren](https://lauren-mccarthy.com/) talks us through some of her many works concerned with smart speakers, machine listening and social relationships in the midst of surveillance, automation, and algorithmic living. We discuss: [LAUREN](https://lauren-mccarthy.com/LAUREN), for which she attempted to become a human version of Alexa, [SOMEONE](https://lauren-mccarthy.com/SOMEONE), which won her the [Prix Ars Electronica 2020 / Interactive Art +](https://ars.electronica.art/homedelivery/en/winners-prix-interactive-art/), and a range of related works and political questions.
+{{% interview "strengers-kennedy.md" %}}
 
-![Interview conducted on 22 September, 2020](audio:https://machinelistening.exposed/library/Lauren%20Lee%20McCarthy/Lauren%20Lee%20McCarthy%20(23)/Lauren%20Lee%20McCarthy%20-%20Lauren%20Lee%20McCarthy.mp3)
+{{% interview "li.md" %}}
 
-## Alex Ahmed
+{{% interview "andre-dao.md" %}}
 
-[Alex](https://scholar.google.com/citations?user=Gc8T8LkAAAAJ&hl=en) talks to us about [Project Spectra](https://github.com/project-spectra), an online, community-based, free and open source software application for transgender voice training. We discuss speech pathology and the politics of pitch, along with the importance of grass-roots led tech projects and community-centred design.
+{{% interview "angie-abdilla.md" %}}
 
-![Interview conducted on 21 September, 2020](audio:https://machinelistening.exposed/library/Alex%20Ahmed/Alex%20Ahmed%20(22)/Alex%20Ahmed%20-%20Alex%20Ahmed.mp3)
-
-## Stefan Maier
-
-[Stefan's](http://stefanmaier.studio/info/) 2018 [dossier on machine listening for Technosphere](https://technosphere-magazine.hkw.de/p/1-WaveNet-On-Machine-and-Machinic-Listening-a2mD8xYCxtsLqoaAnTGUbn) puts the work of artists like George Lewis, Jennifer Walshe, Florian Hecker, and Maryanne Amacher into conversation with Google's wavenet. We talk about these and other works along with Stefan's own compositions which treat machine listening as a prepared instrument, ready to be detourned.
-
-![Interview conducted on 11 September, 2020](audio:https://machinelistening.exposed/library/Stefan%20Maier/Stefan%20Maier%20(21)/Stefan%20Maier%20-%20Stefan%20Maier.mp3)
-
-## Yolande Strengers and Jenny Kennedy
-
-[Yolande](https://research.monash.edu/en/persons/yolande-strengers) and [Jenny](https://www.jennykennedy.net/) provide a “reboot” manifesta in their book [_The Smart Wife: Why Siri, Alexa, and Other Smart Home Devices Need a Feminist Reboot_](https://www.audible.com.au/pd/The-Smart-Wife-Audiobook/1705255280), which lays out their proposals for improving the design and social effects of digital voice assistants, social robots, sex robots, and other AI arriving in the home.
-
-![Interview conducted on 10 September, 2020](audio:https://machinelistening.exposed/library/Yolande%20Strengers/Yolande%20Strengers%20and%20Jenny%20Kennedy%20(10)/Yolande%20Strengers%20and%20Jenny%20Ken%20-%20Yolande%20Strengers.mp3) 
-
-
-## Jùnchéng Billy Lì
-
-[Billy](https://lijuncheng16.github.io/index.html) tells us about his research on "adversarial music", and in particular an attempt to produce a ["Real World Audio Adversary
-Against Wake-word Detection Systems"](https://machinelistening.exposed/library/BROWSE_LIBRARY.html#/book/fac6c1a2-946f-43c4-83f5-e54fd7185c18) for Amazon Alexa.
-
-![Interview conducted on 7 September, 2020](audio:https://machinelistening.exposed/library/Juncheng%20Billy%20Li/Juncheng%20Billy%20Li%20(25)/Juncheng%20Billy%20Li%20-%20Juncheng%20Billy%20Li.mp3)
-
-
-## André Dao
-
-[André](https://andredao.com/) talks to us about [UN Global Pulse](https://www.unglobalpulse.org/), the UN's big data initiative, and in particular one program which "uses machine Learning to analyse radio content in Uganda". We discuss the increasing entanglements of big tech, the UN and human rights discourse more broadly, as well as an emergent "right to be counted". 
-
-![Interview conducted on 4 September, 2020](audio:https://machinelistening.exposed/library/Andre%20Dao/Andre%20Dao%20(26)/Andre%20Dao%20-%20Andre%20Dao.mp3)
-
-## Angie Abdilla
-
-Angie talks to us about [Old Ways, New](https://oldwaysnew.com/), the Indigenous owned and led social enterprise she founded, based on Gadigal land in Redfern, Sydney. We discuss [Decolonising the Digital](http://ojs.decolonising.digital/index.php/decolonising_digital), Country Centered Design, a methodology which applies Indigenous design principles to the development of technologies for places, spaces and experiences, and how this contrasts with the 'placelessness' on which so many machine learning/listening systems are based.
-
-![Interview conducted on 2 September, 2020](audio:https://machinelistening.exposed/_preview/library/Angie%20Abdilla/Angie%20Abdilla%20(31)/Angie%20Abdilla%20-%20Angie%20Abdilla.mp3)
-
-
-## James Parker (w Jasmine Guffond)
-
-This is the first of three radio shows as part of [Jasmine's](https://jasmineguffond.bandcamp.com/album/microphone-permission) guest residency at Noods Radio. It features an interview with [James](https://law.unimelb.edu.au/about/staff/james-parker) about his research on machine listening, this curriculum, the project with Unsound, and a selection of electronic music.
-
-![Interview conducted on 1 September, 2020](audio:https://machinelistening.exposed/library/James%20Parker/James%20Parker%20(30)/James%20Parker%20-%20James%20Parker.mp3)
-
-
-## Vladan Joler
-
-Vladan walks us through [Anatomy of an AI System](https://anatomyof.ai/), his 2018 work with [Kate Crawford](https://www.katecrawford.net/), which diagrams the Amazon Echo as an anatomical map of human labor, data and planetary resources. We talk about the politics of visibility and method as well as Vladan's work with [Share Lab](https://labs.rs/en/), "where indie data punk meets media theory pop to investigate digital rights blues." 
-
-![Interview conducted on 1 September, 2020](audio:https://machinelistening.exposed/library/Vladan%20Joler/Vladan%20Joler%20(part%202)%20(29)/Vladan%20Joler%20(part%202)%20-%20Vladan%20Joler.mp3)
+{{% interview "parker-guffond.md" %}}
 
+{{% interview "vladan-joler.md" %}}
 
 ## Halcyon Lawrence
 

content/topic/lessons-in-how-not-to-be-heard.md

@@ -1,6 +1,7 @@
 ---
 title: "Lessons in how (not) to be heard"
 has_experiments: ["adversarial-music.md", "ad-versarials.md"]
+
 sessions:
   [
     "unsound-2020-session-2.md",
@@ -45,9 +46,10 @@ Enjoy being listened to; get your coronavirus diagnosis over the phone; get your
 
 ![ML](static/images/01-1280x706.gif)
 
-# Bibliography
-[^Forensic]: LIBRARY Forensic Architecture, Violence at the Threshold of Detectability (2017),p27.
-[^forensic listening]: ![](bib:731924ff-868d-4692-a184-b28e335e1871); James Parker, "Forensic Listening in Lawrence Abu Hamdan’s Saydnaya (the missing 19dB)" *Index* (2020).
+# Resources
+[^Forensic]: Eyal Weisman, 
+_Forensic Architecture, Violence at the Threshold of Detectability_ (Zone, 2017),p27.
+[^forensic listening]: ![](bib:731924ff-868d-4692-a184-b28e335e1871); James Parker, ["Forensic Listening in Lawrence Abu Hamdan’s Saydnaya (the missing 19dB)"](http://index-journal.org/media/pages/issues/law/part-2-lacunae/forensic-listeningin-lawrence-abu-hamdans-saydnaya-the-missing-19db-by-james-parker/1548323910-1602118907/james-parker.pdf) *Index* (2020) HTTPS://DOI.ORG/10.38030/INDEX-JOURNAL.2020.2.7 
 [^Abdilla]: Interview with Angie Abdilla, founder of [Old Ways, New](https://oldwaysnew.com/) on 2 September 2020
 [^Ahmed]: LIBRARY Alex Ahmed, 'Online Community-based Design of Free and Open Source Software for Transgender Voice Training' (2020); ![](bib:2a16a513-96d8-4d68-bd70-23f74a71a609)
 [^McCarthy]: Interview with [Lauren McCarthy](https://lauren-mccarthy.com/LAUREN) recorded on 22 September 2020.

content/topic/listening-with-the-pandemic.md

@@ -57,14 +57,14 @@ Big tech knows you can get away with things in assisted living facilities that y
 
 For such an ambient sensing environment to work, this very environment must be designed and shaped with embedded cameras and microphones in mind. Every room becomes a studio. Background noise must be minimized to make objects and sounds a little more legible. And we know that such environmental design doesn't stop at objects and spaces: it reshapes our own patterns of speaking and living as we learn to enunciate with a cadence, accent and tone that an algorithm can understand.
 
-# Bibliography
+# Resources
 
 [^Mattern]: Shannon Mattern, ['Urban Auscultation; or, Perceiving the Action of the Heart' *Places* (2020)](https://placesjournal.org/article/urban-auscultation-or-perceiving-the-action-of-the-heart/)
-[^Schuller et al]: LIBRARY, Schuller et al, COVID-19 and Computer Audition: An Overview on What Speech & Sound Analysis Could Contribute in the SARS-CoV-2 Corona Crisis (2020)
-[^Szendy, Vetter]: LIBRARY, Peter Szendy, All Ears; Grant Vetter, The Architecture of Control
-[^Sterne, Rice]: LIBRARY, Jonathan Sterne, “Mediate Auscultation, the Stethoscope and the ‘Autopsy of the Living’: Medicine’s Acoustic Culture,” Journal of Medical Humanities 2, (June, 2001): 115–36; Tom Rice “Learning to Listen: Auscultation and the Transmission of Auditory Knowledge,” Journal of the Royal Anthropological Institute 16, no. 1 (2010): 41– 61.
-[^Goldenfein]: Jake Goldenfein, Monitoring Laws; Ken Alder, “A Social History of Untruth: Lie Detection and Trust in Twentieth-Century America,” Representations 80, no. 1 (November 1, 2002): 1–33, https://doi.org/10.1525/rep.2002.80.1.1.
-[^Wark]: LIBRARY, McKenzie Wark, Capitalism is Dead
-[^McQuillan]: LIBRARY, Dan McQuillan, Data Science as Machinic Neoplatonism
+[^Schuller et al]: Schuller et al, [COVID-19 and Computer Audition: An Overview on What Speech & Sound Analysis Could Contribute in the SARS-CoV-2 Corona Crisis](https://arxiv.org/abs/2003.11117) (2020)
+[^Szendy, Vetter]: ![](bib:d6b5c725-9964-4218-9157-b3d6fd7ca62a); Grant Vetter, _The Architecture of Control_ (Zero, 2012)
+[^Sterne, Rice]: Jonathan Sterne, “Mediate Auscultation, the Stethoscope and the ‘Autopsy of the Living’: Medicine’s Acoustic Culture,” Journal of Medical Humanities 2, (June, 2001): 115–36; Tom Rice “Learning to Listen: Auscultation and the Transmission of Auditory Knowledge,” Journal of the Royal Anthropological Institute 16, no. 1 (2010): 41– 61.
+[^Goldenfein]: ![](bib:6e8f7c36-d251-4a07-ac5d-0b938c5f5fee); Ken Alder, “A Social History of Untruth: Lie Detection and Trust in Twentieth-Century America,” Representations 80, no. 1 (November 1, 2002): 1–33, https://doi.org/10.1525/rep.2002.80.1.1.
+[^Wark]: McKenzie Wark, _Capital is Dead: Is this something worse?_ (Verso, 2019)
+[^McQuillan]: ![](bib:491e5855-378e-4882-8f6e-c0f1d1099fe3)
 [^andrejevic]: Interview with Mark Andrejevic recorded on 21 August, 2020
-[^Abu Hamdan]: Interview with Lawrence Abu Hamdan recorded on 13 September, 2020. See also Abu Hamdan, "H[gun shot]ow c[gun shot]an I f[gun shot]orget?" (2016), http://lawrenceabuhamdan.com/blog/2016/3/7/hgun-shotow-cgun-shotan-i-fgun-shotorget
+[^Abu Hamdan]: Interview with Lawrence Abu Hamdan recorded on 13 September, 2020, publication forthcoming 2021. See also [Abu Hamdan, "H[gun shot]ow c[gun shot]an I f[gun shot]orget?"](http://lawrenceabuhamdan.com/blog/2016/3/7/hgun-shotow-cgun-shotan-i-fgun-shotorget) (2016) 

content/transcript/abdilla.md

@@ -0,0 +1,84 @@
+---
+title: "Angie Abdilla"
+status: "Auto-transcribed by reduct.video with minor edits by James Parker"
+---
+
+Joel Stern (00:00:00) - It would be great if you, if you could introduce yourself and give a bit of background to the work that you do and how you came to sort of think about advanced technologies through it, through the prism of, of indigenous culture and knowledge. 
+
+Angie Abdilla (00:00:14) - Uh, so, um, so I'm a Palawa woman. So my, uh, my mob come from the North Eastern parts of Tasmania, uh, also, um, we have a more recent history in the Northwest of Tasmania, but, um, uh, so we I've been living here in Sydney for most of my adult life, but I go back home regularly. And I guess, um, the reason why I became quite, uh, intrigued with technology and, and started to develop a much closer relationship with it and a working relationship is because as a filmmaker for many years, I was, uh, seeing how the way we tell stories. So originally, you know, in the, from the rectangle know cinema to the square TV, and then, and then this thing, internet and content transmedia storytelling and, and con you know, our films and media becoming content really, I guess, started me thinking about how these, um, these transitions in the way we tell stories and those shape and format of those stories was changing.  
+
+Angie Abdilla (00:01:27) - So radically quickly in the broader spectrum of time, those transitions between cinema and broadcast media were actually quite pronounced, you know, they weren't, they happened over a much longer period of time. And so I was really interested in how digital technology was creating these fundamental shifts in ways that were really, um, exciting for me. Like I started seeing this realm of possibility. I remember the very first website that I, that I came across was by this Russian artist years and years ago when I was, and there was, there was no way in, you had to, it was almost like a galaxy, which, um, you had to find the right star constellation to enter into this wormhole. It was really great. And then, you know, once you're in, it was, it was this incredible world of discovery. And so, you know, of course it's a far cry to the way that, um, web platforms are designed these days. 
+
+Angie Abdilla (00:02:04) - But what I saw was this incredible opportunity to, to imagine a different way that we could inhabit, uh, a liminal space, a liminal space to connect and to share stories and to share ideas and experiences. And so it was kind of during that time as well, like, you know, there's between like 20, 25 years ago that, um, the very first, I guess, tests around with around VR were also being explored and it wasn't cold VR back then. Uh, um, I do remember this, you know, coming from an arts background and, uh, really interested in installation art and video art, how these, um, the different ways we shaped story, um, through these different platforms and devices could create an immersive experience. And so that was really, you know, what like that seeded this interest. And so then I was a film director and writer and producer for many years, and I was always frustrated at how limited and contained it was, even though I love the craft and form. 
+
+Angie Abdilla (00:03:39) - It was still compared to, um, the experience of making him of the capacity of immersive contemporary art. It was very limited, I thought. And so, so when I started seeing these different ways that, that the film industry was shifting, I started thinking, well, we have to, I have to beam to think more deeply about this and, and to cause this there's an important shift happening, and I need to be part of that. So, yeah, that's really how the company came about. I mean, there was a whole bunch of other different projects that kind of led me on this journey. Um, one of which was a, a role that I was doing many years ago now in, for the national center for indigenous excellence. Whereas leading this initiative called the indigenous digital excellence initiative. And through that role, I was fortunate enough to explore what the, where the, what I considered to be the most, um, important new technologies that could be important for our communities. And so back then I was thinking about, well, I was really interested in robotics, 3d printing and gamification. And so I developed these prototype workshops.
+
+Angie Abdilla (00:04:50) - For, um, for young kids, for Aboriginal kids to introduce this concept of code and how we work with code and these different forms. And so the most successful out of those workshops was the robotics workshops, which then was the springboard for writing a research paper with fiber culture journal was a special edition on robotics and that, um, was a peer reviewed paper. And it was really exploring. I mean, it was a sort of, I guess, a summary of what the prototype workshop was, but really it was a, it was an opportunity to delve deeper into what are indigenous knowledge systems and pattern thinking. And how can those, how can we think about the possibilities of new emerging technologies beyond this access and participation in gender that most people kind of foist on Aboriginal people and think more about how our knowledge systems, our cultural knowledges and knowledge systems can inform the conception, design and development of technology. 
+
+Angie Abdilla (00:05:55) - That's really what was the beginning of the company. Um, but that's really, you know, the, the, the company itself was founded by my uncle and I say, he's passed away last year, but he was a total futurist. He was a law man initiated law man. And, um, but a total futurist; not at all scared of technology, like just excited by the possibilities. Like, you know, he's always talking about quantum physics and there, and its relationship to indigenous knowledges and how, how relevant it was and how, you know, he was, um, he was a poet, he was many things, but he had this incredible ability to see different possibilities. And he was really the driving force, uh, to guiding the formation of the company. Really.  
+
+Joel Stern (00:06:44) - I read his biography on the website and it sounds like a remarkable man and yeah, you are very lucky to have had, uh, a mentor like that. Um, I wonder if you could, um, talk a little bit about country centered design and pattern thinking, um, you already brought up, brought up pattern thinking, um, and in, in the last answer and they're two sort of concepts that, you know, inform the methodology really strongly, um, how do they kind of operate in your work and, and what are the imperatives? 
+
+Angie Abdilla (00:07:22) - Mm, so pattern thinking, um, as I wrote about in that first research paper, um, is a way for us to understand, to, I guess, um, articulate the interrelationship and interconnections of all things. And, you know, this is very relevant to, I guess, physics and quantum physics. So if you think about Australia as a living organism and massive, um, is the continent is held together by this, this network of some lines that intersect and interconnect. And they all come from this one central place, which is, um, some say the womb, some say the navel, and that's all, there's some lines come out from that place and they hold the continent together. In a sense, what I describe it as like the central nervous system, your country, when you look at that as an entire network within all of that, there's a, on a macro level, there's a way of understanding the, um, the didactic nature of the con of the continent. 
+
+Angie Abdilla (00:08:31) - So there's different purpose for those different, the different, um, areas of this very vast land it's dry and arid parts. We've got salt, water, fresh water. We have the, the top parts of the continent, uh, um, rainforest and actually the lungs of the continent and the bottom Tasmania and kangaroo Island like the feet. And then you have the womb in the middle, you know, so it it's a living organism in itself in its own, right? On a macro level, on a microscopic level, the, those interrelationships and interconnections within that, within the continent go down to, you know, the, my relationship to, from where I'm sitting right now to the, to the Harbor, to the mountains and, and otherwise, you know, like it's a way of, um, we're all interrelated and interconnected, you know? And so it's much more, I guess, obvious for, I guess, another way of understanding that is through the kinship system. And it's much more obvious for mainland, um, the mainland continent, because it's, those connections are very, still very much still connected. Whereas we've, you know, for in Tasmania, we've been disconnected physically.
+
+Angie Abdilla (00:09:50) - For 12,000 years, but also disconnected through the decimation of our culture by just in the last 200 years. I mean, it is only the last 200 years and those connections are still there, but it is a reality. So the, so there's different ways of understanding, pattern thinking. And what I talked about in that paper was how to the two different, like, I guess from, um, anthropological kind of perspective, it's, you know, think about epistemologies and ontologies in some ways, you know, indigenous ways of being, knowing, seeing, sensing, doing it's all kind of the, it kind of collapses in, on each other as the one thing when you're able to be in that, in that state of connection. So there's pattern thinking and what I, what my uncle also used to talk about was pattern recognition. And that for me is the it's like the mother tongue it's for me, you know, um, have grown up in a post-colonial society.  
+
+Angie Abdilla (00:11:00) - So I'm, I'm learning those cultural practices again, but there are still a lot of people that still have that really strong in a way, you know, like that Mo the mother tongue it's like the mother tongue of seeing pattern recognition is the sort of saying sensing, doing the mother tongue of that. So I guess, um, all of that work and that quite deep sort of philosophical kind of framing really was foundational for, for understanding that there was, uh, there's a vast amount of knowledge that, uh, that can support the conceptual design and development of technologies that have these principles of caring for country and caring for kin and, and law. They're really kind of deep, quite, um, intricate law that, that holds different communities together in different country together. So what I saw was is this very obvious pathway that we could start working with that law, working with those knowledges as our people have done for thousands of years into design different solutions to complex problems, that's, what's always happened. And so why should we not be doing, continuing to do that, but with new emerging technologies? 
+
+Sean Dockray (00:12:30) - Yeah, I think that, um, that relational thinking that's sort of at the heart of, uh, the knowledge that you're describing is, is clearly something that like science and technology kind of developed in a kind of Euro European context is like catching up too in so many ways, because that whole scientific project has been based in kind of separating and analyzing and looking at things in isolation. And only now, only recently, I feel like scientists and engineers are sort of realizing, Oh, in order to do anything about the complex problems, we have to look at relationships between things, you know, so as if it's a big and novel discovery. And so, but this relational thinking as you're describing is just part of indigenous style, culture and knowledge in order to be able to apply that to the kind of conception and design part of projects, which you emphasized at the beginning. And I just see it's so important to get in at the stage of conception and not sort of like cleaning up the problem or operating in a tiny little silo, but actually being able to, to conceive, to, to operate at that place where you have access to lots of different, um, parts of the problem. Has it been hard for you to make that kind of like, um, argument or like, how do you go about being able to get to operate at that stage of conception? Do you know, does that question make sense? 
+
+Angie Abdilla (00:13:59) - So the company was, has two arms. We are a consultancy and we generate a rare proprietary limited company and a social enterprise. So the consultancy arm works within the built environment and the cultural sector, delivering services in a lot of ways to working on large infrastructure projects, uh, very much with, you know, the engineering and the engineering and technologies and how they reside within place. It's a place making, and I would say the, the shaping of those projects. So that's on one level, those profits then support the research and development works. And so the whole formation of the indigenous protocols and artificial intelligence working group was funded by all of those projects, along with some other funders. But prior to that,
+
+Angie Abdilla (00:14:52) - The profits also supported a book. We wrote called Decolonizing the Digital: Technology as Cultural Practice. Prior to that, it was also the funding also supported the, um, the very first research paper, which also was kind of foundational to the establishment of the company. It was a little bit sort of chicken and egg in some ways, but I guess though, that's important to just note because the, yes, it is hard and it's not, it's not like this has just happened overnight. We've spent years and years and years setting up the foundations to enable us the position that we're in right now, where we're able to work with various different partners who have capacity to, to work with S w work with us on that level. But it takes time and it takes trust and it takes deep working relationships that don't, then it just doesn't happen overnight.  
+
+Angie Abdilla (00:15:49) - So some of those projects, like for example, um, we were working with the Barangaroo delivery authority, which changed over to net is now known as infrastructure, new South Wales and what we were, one of the projects we've worked on and it was called the big sky. And essentially what it is, is a large scale public art installation. But what it does is, um, it's still in, um, it's kind of on hold at the moment because there's a whole heap of work being still being signed off on underneath our site. So outside, which is kind of above ground. Um, there's a whole site called the cutaway on the internal cavity of the Barangaroo Headland is well, there's a massive cavity. So because of the planning, ours is on hold, but essentially it's kind of ready to go. What it is is it's a experiential way of understanding the interconnection and interrelationship of the sky and the constellations within the best guidance, the star stories and its relationship with the Songlines on.  
+
+Angie Abdilla (00:16:57) - So the, the whale dreaming and the sea, so country is those three elements always: it's the water. So whether it's fresh water or salt, water is sky and it's the land, it's all those three things. And we are part of it and it is part of us. So what we were able to do in that project was to design an, a site and an experience through connecting with those diff those costs on those core Songlines. Um, so the whale dreaming, but then in the sky, the seven sisters. So the seven sisters spans from the West across the whole of the continent. And likewise, the whale dreaming spans from Tasmania all the way up past, even this continent and on like, there's lots of different indigenous moms that have whale dreaming. So it was really those two really big important Songlines. And when I, that the star constellations are the mirror images of those Songlines.  
+
+Angie Abdilla (00:17:56) - So when you look up and see those different constellations in the sky, they are mirror images of what the, where those song lines are and how they learn. So the site itself, um, creates an opportunity for visitors to, to immerse into that, can that, uh, relationship of the whale and the stars through these different, um, technologies embedded within the site. So the methodology we employ for all of our projects, country centered design, what it does, is it frames a process there's sort of guiding principles. Um, it's not prescriptive. It's not, um, because it can't be, and I'm always really reticent to documenting it because as soon as, you know, we do, it becomes problematic, but essentially what it is is there's four core processes, which is culture, research strategy, and tech to culture. Some of the activities always start with working with custodians, knowing whose country you're on, and how do you develop a relationship, a deep relationship with that country you're working with, what's the, you know, one of the ways of mapping it and how do you, um, how do you know it then it's, um, developing up a rich, uh, knowledge base of those, of those cultural practices and protocols and right, and rituals that come from that country, that clan group and how those different they're different community members you're working with can support that process.  
+
+Angie Abdilla (00:19:30) - That then helps us to inform the research component. So the traditional knowledge and that other associated researchers that can come, can support that piece, which then informs the strategy. So how we design out the solution and we do all sorts of various different, you see all sorts of different techniques and futuring and, um, interaction design, and all sorts of different modalities are used.
+
+Angie Abdilla (00:19:56) - Two then finally, only once we've gone through that quite intense process, we get to the, what are the right technologies for the context? What is the cultural foundations that we've established to develop a conceptual framework and the creative strategic design interventions that then can inform the right technologies for the context. And so in that, for this project, it was, um, it was predominantly a sound experience yet. And these, there is a massive dome that has thousands of led lights programmed inside the dome to help you navigate where the stars are, but predominantly there's different sound technologies that, um, like bone conducing technologies that and variety of other different, couple of other different, different things that help channel these different stories sounds and experiences through you to connect to these different elements. So the dreaming stories, the soundscapes, and then the other contemporary stories. So that's more on the aspect of, and then on the other end of the spectrum, we've done projects like with, um, you know, on a precinct level. 
+
+Angie Abdilla (00:21:11) - So we've done quite a few of those where we're designing entire precincts, where often there'll be a train station at the heart of it there'll be the, as a public domain element, there's retail strategy. There's a often there's, um, residential, private and social sometimes. Um, and otherwise, so these have been complex X, big precincts need to be designed. And so we often are working on those pre large projects from the early conception part where we think about what is the origins of this place. So a recent example, we've been working on a project, a master plan for Macquarie park, which is one of Wallumettagal country. So that Klan group, uh, the Wallumettagal is the, um, is the people of the black snapper fish. And so the two core dreamings, uh, the whale, uh, sorry, the, the eel and the black snapper, and they, uh, hold, they kind of have the main boundary lines for that particular clan group.  
+
+Angie Abdilla (00:22:17) - And so we designed a whole different array of strategies for how the, the didactic cultural practices can inform the different strategies for that entire master plan. So, um, the fine grain network. So how does the transport network work? How do we design that around the origins of how the river, their network of the riverways was used? Yeah, there's a whole bunch of different examples. I can give you on that, but that's, I guess the, sort of the, the two ends of the spectrum of how we work in these different projects and how that methodology country centered design is. And it enables us the process that ensures that we're always developing a really, um, strong integrity within the cultural foundations for the conceptual framing, the strategic design, and then the technology intervention. 
+
+James Parker (00:23:13) - That's so interesting, Angie. Um, and as you're speaking, I'm just thinking everything you've said sounds like, you know, you'd never hear it from a Google exec, you know, or an Apple exec, uh, and that, you know, something like, you know, Oh, you know, obviously sadly, um, but that, you know, w we've been interested in Machine Listening as, you know, it's sort of embedding itself into our homes and our, and our cities and, uh, and so on. And part of the conceit of it is, you know, a kind of placelessness that it can just be rolled out anywhere. And that one home is the same as another, and all sorts of things happen then, you know, because you don't design for accent or language, and you're certainly not designing for country when you design for something like Alexa. Yeah. I just wondered if I don't know how, I don't know how, I don't know what would be a way in to comment on that. 
+
+James Parker (00:24:13) - Um, because it just seems like a complete that, that the two approaches are completely anathema that, you know, that big Tex business model is big because it's kind of colonial or Imperial is precisely not interested in those questions that you've been talking about. But, but yet when you speak, you sound, you speak very optimistically about technology and you talk about your uncle being a futurist and so on. And I wondered if you could comment on how to hold those different things together. They're kind of the place and listeners or country listeners of so much big tech thinking and what, what interests you? 
+
+Angie Abdilla (00:24:52) - Oh, well, I think you're bang on, I mean, what you're highlighting as well as the, this, um, the normative language that comes with.
+
+Angie Abdilla (00:25:01) - Bodiless stateless knowledge that drives these, um, I guess the it's Imperial technology drive, you know that, yeah. I mean, I, I don't know. I think that the, yeah, I'm really optimistic. I think that there's, what's, what's really apparent to me is that there's so many different organizations right now. Like we're seeing within COVID crumbling because they're, they're not sustainable the actual systems themselves, the business models that they operate within, uh, completely unsustainable and they're, and they're, they're really suffering and now's the time to let them crumble them. I say, I mean, there's, you know, it's a really tricky, it's a pretty hard line because of course there's people's lives that, that are entwined within those big systems. But, um, what's, I think what's the, at the heart of the, the unsustainability of those systems, whether it be, um, a corporate organization or a technology platform or something else is how devoid of culture they are and the cultural protocols that shaped their identity and shaped who they are, how they operate and the, the, and I mean, they're not devoid of culture. 
+
+Angie Abdilla (00:26:32) - They're just the culture within them is really toxic. And so that's why I think a lot of these large entities or systems and, or systems, uh, failing, because there's no integrity that's holding them together. So I know that's really kind of, that might be a little bit esoteric, but, um, yeah, that's what I think is different to the work that we do is that where, and I don't want to sound really Ernest, but that's, that's what I know has sustained our people and how this country has nurtured us for so many thousands of generations is because of our culture. It's the culture and it's those protocols within our, within our law and within country that has sustained us. And so that those are the protocols that inform everything, how we have designed these technologies over thousands of years, like the boomerang fish traps, spinifex resonant worlds, you know, there's so many different examples of those different technologies that have been designed with such that it absolutely imbalance and in harmony and, uh, a symbiotic relationship with, with country, you know, so there's this, you, you don't ever take more than what you can, what you need, but there are times though there are, there are needs for, there are reasons to take more.  
+
+Angie Abdilla (00:28:17) - And so there are different protocols around that. Yeah. So the, I mean the actual concept that a lot of environmentalist's have have supported over time. Doesn't, doesn't really align with indigenous ways of being, and thinking and knowing, and seeing and doing, we don't lock country up as thing that needs protecting there's a, it's far deeper and far more complex than that. And so, you know, we work with it. It gives us as much as we in. So yeah, it's a, it's a relationship. So it's the same thing with technology, you know, and I think, you know, you talked about something earlier about, um, how do we develop meaningful relationships with technology and in particular, when thinking about Machine Listening? No, my uncle once said to me that, you know, we've gotta be really careful about the types of, um, AI that we're developing, because where w w we're potentially going to be creating another enslaved race. 
+
+Angie Abdilla (00:29:25) - Cause, you know, if you think about, um, AI as a, a lot of, a lot of indigenous peoples actually do believe that there is sentience within, within maybe deep learning and general AI, you know, there's, it can be, that can be, you know, debated for awhile. But, um, but I do believe that there's, that we're heading down a road where there, there is the capacity for these technologies to have a different light force. And so then therefore the question needs to be asked, like, what, what are we creating if we are creating another life, another.
+
+Angie Abdilla (00:30:03) - Life force. Then we've got to be really mindful about, you know, the, the, the ethics around, you know, those different types of technologies and their role and their place and how we as humans relate to them. 
+
+Joel Stern (00:30:18) - Um, thanks Angie. Um, for those thoughts, um, one of the essays that you mentioned when we chatted on the phone last week, um, the, um, Making Kin with Machines was re was really, um, stimulating to, to, to read that into, to kind of go deeper into those questions of, um, human Machine interrelations and the sort of ethical implications of that I'm aware. We've just got a few minutes left. So I wanted to ask a, kind of a more speculative, um, kind of question about Listening technologies and, you know, if you were to sort of, um, sit down and think about Listening technologies or indeed, you know, Aboriginal technologies of Listening that have a rich set of cultural protocols already existing, but thinking about technologies of Listening through that prism of countries and to design, you know, what, what would those technologies ideally do? What, what would they produce? Like how might they be used? And I think that's a very big question, but, um, from, from our perspective, if we could get you to think that through then it would be a really amazing contribution to this project and, and something that, you know, would be really special.  
+
+Angie Abdilla (00:31:44) - So if you haven't heard of the message stick 
+
+Joel Stern (00:31:47) - A little bit better. Yeah. Could, could you, 
+
+Angie Abdilla (00:31:50) - Uh, so the message stick is often, um, it's a, it's a stick and it's got lots of different, uh, symbols carved within it. And it's used as a, as a communication tool. So it's, um, it's embodied knowledge, but the, so that it's often, you know, it was used for various different communication reasons, but, um, but the, the message is also used as a, as a way of, of holding space as well. So when yarning circle, there are different ways you can facilitate and hold the yarning circle. And those protocols are, are really important for, for establishing, I guess, uh, um, deep Listening. So, so in, in those cultural practices, there are, there are design opportunities for different emerging technologies. And if we look at the benefits of, you know, what comes from what comes out of the, of a yarning circle, they're amazing with that when they're facilitated really, really well.  
+
+Angie Abdilla (00:33:02) - And what they enable is for people to speak deeply, um, in ways that I've never really seen before actually. But there's also another element that goes with that yarning circle. That what, what we do is we teach people how to make string for, for weaving. So there's something in the act of actually doing something with your fingers, like knitting or wood, there's many different things you can do that very active using your hands while in a yard, while in a circle sharing is kind of creates a, a different way that you can open, open up. And so all of these cultural practices can help us to think about what are the, what are the different ways that we can think about the design, the consent, the intent, and the need, and the can, the conception and design of these new, um, Listening technologies. I think that there's incredible opportunities to, to embrace AI for those different reasons, but also, you know, too, we need to also really critically think about, well, how, what form of AI, you know, there's this assumption that it's Machine learning or deep learning on general AI and, and that's pretty much, you know, those are, that's what we're, that's what we have.  
+
+Angie Abdilla (00:34:25) - And yeah, that's currently what we have, but it's not to say that there's, you know, we should start thinking about what as indigenous peoples. And this is what this whole work is really about. And the indigenous protocols and artificial intelligence work is what is the AI that for, that we want to design for the future and how, what are the protocols that embody it? And what are the ways we understand? What are the protocols that inform the way we, we care tech for knowledge and how are they being replicated within a synthetic environment and for, and how do we.
+
+Angie Abdilla (00:35:07) - How, yeah. What are we saying yes. To when we're allowing that knowledge to become data 
+
+Joel Stern (00:35:14) - The indigenous protocols, artificial intelligence paper, um, is, is amazing. Um, it's really an incredible resource. There was a, um, quote, I think Dr. Noelani Arista that really stood out where she said, um, the key will be to ensure that the intellectual architecture preserved orally and textually by our ancestors helps shape the computational architecture of, of our digital technologies. And I thought that, you know, maybe sh she was hinting at, um, you know, like language and cultural preservation that could be, um, enabled by AI. And in fact, one of the people we spoke to at Mozilla who was working at Mozilla in, in, um, voice recognition, um, pointed to this as a specific use that, um, voice assistance might be able to assist in language preservation and conservation sort of, do you have a sense of what the kind of opportunities and sort of, um, challenges obstacles might be for a project like that?  
+
+Angie Abdilla (00:36:18) - Yeah. Yeah. Well, I think on one hand, I think it's so important because, you know, coming from Tasmania, we had all of our languages were decimated. And so right now there's a language revitalization program happening piecing together the nine different languages into one, one dialect. And it's not perfect. It's not, you know, we've only got journals to go from. And so imagine if there was a different, um, archive, however, um, what's really important is the protocols that underpin knowledge transmission in the first place. So there's cultural practices and protocols that inform how you, you know, when you're ready for that knowledge is super important. So, I mean, it's kind of different with language when you're learning it again as a second, not as your mother tongue, but as a second language in that scenario, I think, yes, we should. We just need to preserve our languages, but we also need to, it's a living culture, so we need to be practicing them. 
+
+Angie Abdilla (00:37:24) - Um, so it's not, so the pre preservation is always a bit sticky for me, you know, um, when you have a living culture, it's about the actual act of it. It's the practicing of, and so technology right now often is about, um, cultural preservation or protection instead of cultural practice. So I think that's the, one of the key differences with the book that we wrote decolonizing, the digital technology as cultural practice. Once again, it's, you know, it's not the same agenda of, um, access participation, preservation. One of that, it's very, it's a very different agenda and it's not like I'm, poo-pooing the whole, like all of it, but it's, but we need to, we need to invest in another agenda equally as much, which is about agency and autonomy and the systemic change that can be created by, by developing different types of systems and technologies that come from, from those, the systems that are inherent to this country and the co and the cultural practices that are part of those. So the actual designing and developing of those technologies becomes part of the culture.

content/transcript/ahmed.md

@@ -0,0 +1,114 @@
+---
+title: "Alex Ahmed"
+status: "Auto-transcribed by reduct.video"
+---
+
+James Parker (00:00:00) - Could you maybe kick things off just by telling us a little bit about yourself, your work and it just a bit of a general introduction. 
+
+Alex Ahmed (00:00:06) - Yeah. So my name is Alex Ahmed, I just finished my PhD in health informatics at Northeastern university. That's located in Boston, Massachusetts in the us and in my dissertation work was, um, it started kind of with the goal of exploring voice training technology. I was coming out at the time as I was trans. And, um, I kind of saw this as like a project that would make me feel like that I was doing something worthwhile. Um, and it was like originating directly from my experience. And that was back in 2014. So that's when I started working on this. And about six years later, I had a prototype out for, um, for an app that, um, was created by, by me and also a group of other people who came to be called project spectra. And this was, um, a, a community of about 20 people, um, that worked together for like a year and a half to make this a reality. So it was really grounded in like interviews that I conducted myself with other trans people about waste and then, um, kind of taking those experiences and translating them into a potential design. So there was kind of that grounding and lived experience. And then using that to like, try to figure out what would be useful, but also examining what's currently out there from a sort of critical, um, queer theory lens. So those were like the three main components of the project. And yeah, so hearing him,  
+
+James Parker (00:01:56) - It's an amazing project. Can you, can you say something about the, sort of the nature of the intervention, that project spectrum makes? Like what, what was the field of voice training apps like, or perhaps it even its history beyond, you know, the apps and, and so what, what is it that project spectra sort of attempts to do in that context? 
+Alex Ahmed (00:02:15) - Yeah. So, um, voice training as a clinic, as a clinical field for specifically for trans people is relatively young. I don't think it's necessarily like the case that no trans people went to speech therapists, like with the goal of feminizing masculinizing, or just generally modifying their, their voice before, like, you know, the nineties or the early two thousands. But that is when the first like clinical handbook for speech therapists working with trans clients comes out. Um, and then that text has been revised and updated several times since then. But voice training, like from that clinical standpoint with practices geared specifically towards trans people is pretty new. And the intervention that I was trying to pursue was on a few different levels. I think number one is the, like the way that clinical understandings of trans people are structured by these normative ideas. Um, and this goes back a long time. 
+
+Alex Ahmed (00:03:32) - So even outside the realm of voice it was coming. Yeah. The practice for transgender people, um, who were trying to obtain hormones or gender transition treatments of various kinds, um, had to go through a lot of barriers on their way. And a lot of those barriers came from clinical and medical institutions. So trans people would need to, um, to sort of prove that they were sufficiently trans I guess, and that, that happened through a lot of different ways. So one was, you had to kind of fit the profile. Um, and that profile was kind of like, Oh, you had, um, you had thought about like being, you know, a woman or a man, like from an early age, you played with dolls, you played with cars when everyone else it wasn't like that kind of thing. So there's this narrative, this life transit, um, this life trajectory that, um, people were expected to, um, to adhere to. And if you didn't, then it was strange that like, that may be suggested that, Oh, it actually, it, you're not transsexual. You may think you are actually just a sexual fetish or it's just a, it's a phase or it's, um, it's something else. So this actually happened to me, um, when I went to, um, a therapist for the first time to sort of talk through these issues. And that was only in 2012.
+
+Alex Ahmed (00:05:04) - So not too long ago. And of course, you know, that person, I went to isn't representative necessarily of like all clinicians, but it was standard. I think enough that like, unless you were in something very progressive, uh, leaning that in like a clinical center that was more progressive than you could expect that. So these like normative ideas were sort of structured into the very act action of like transitioning. And so that included a history and also like the present. So you had to like prove your transness through what they call the real life experience. So you had to essentially present, look, act as female or male or whatever, you know, you were attempting to pursue transition, but you had to do that without any, any hormones or any, like, you know, if you wanted to have like any sort of like surgeries or anything like that, like that you couldn't do until you had spent a certain amount of time, like a few, you know, a month, a year, whatever, as this real life experience, you know, as the gender you claim to be, and this was sort of like you had to prove it. 
+
+Alex Ahmed (00:06:24) - So this kind of extending also into the voice world, because voice is just one of those things that you had to perform in order to stay sufficiently or convincingly state the claim to like a gender. And so that's kind of, um, partly where, you know, the clinical aspect comes into it and this isn't a claim that like no trans people ever went to a clinical voice therapist, like independently it's that, you know, independent of all of this, like institutionally like mandated stuff, like went to a therapist and said, I want to do this. Um, because certainly they did. And certainly they do, we do, but this is sort of like the middle, you do the social milieu or how this is happening, you know, there's no really there's no world in which like a trans person is not thinking about these things too. It's both, what do I want and what is also expected to be? 
+
+Alex Ahmed (00:07:24) - So for voice we were seeing in like just the clinical world and also in these apps, um, some of which I looked into like, um, you know, early on, um, were like presenting them, presenting a certain idea of like gender of femininity and masculinity that was very stereotypical and very like narrow. And I can show you again, those apps are that I just showed you earlier, but this is kind of like what we were trying to work against. So it's the technology design resulting from this, like these like social pressures and also the social pressures themselves, which, um, which were structuring the designs and technology in the first place. 
+
+James Parker (00:08:13) - Could you maybe describe, um, some of those sort of normative for narrow features, um, just as a way into describing the difference that project spectra is and for, um, you know, uh, I think, um, perhaps if it's possible to do it orally rather than visually for the purposes of this interview, that would help. 
+
+Alex Ahmed (00:08:35) - Yeah. So, um, the way that voice is presented and the way that gender is presented and, um, in voice training apps kind of assumes a lot of things, not always, but often assumes a lot of things about, um, what you're trying to do with your voice. So in one app it assumes that, or it tells you that in order to be sufficiently feminine, you need to be able to hold an 83 pitch. So 220 Hertz. And so the first exercise, the app gives you is a tuner that if you're on a three it's green, and if you're not it's red and it's sort of, it's designed kind of sorta like a target. And if you're off by a little, it turns red. So by definition, it's like narrowing, it's like structuring, like this is where you began, you began. And from there, it, it, it just continues, right? So it's like, do you, are you like being sufficiently breathy? And like your tone is your sentences elevating at the end, like as a sort of up, um, up speak is con is the common word for it? Is that like, you sort of talk like this and that's what you're trained on. So you're sort of meant to mimic what the teacher is doing. So this is very interesting in that a lot of.
+
+Alex Ahmed (00:10:08) - A lot of apps don't do this. They don't usually show a human face and like, have you mimic a human? Um, so like think of like pretty much any app you've ever used, right? Like you never see a person there or people don't talk to you, like maybe you see an in an embedded video, but like in some of these apps that are for voice training, the image of the therapist is very central. And that kind of gives this impression that like, okay, this is my ambassador. Like this is the person who was welcoming me into the gender that I want to be. And so has the knowledge, the credibility, the history, the expertise that I need to copy and to, to achieve my, my gender, my gender goals. That's how these apps are, um, explaining themselves or presenting themselves to be user. So this is really problematic because most, if not all transgender speech therapists that I've, or sorry, not transgender, the speech therapist, speech therapist, who is to like attend to transgender clients, um, they're all cis-gender women and they're all white. 
+
+Alex Ahmed (00:11:25) - That's the sort of general rule that I've seen. I've never seen a speech therapist that isn't one of those. And so when you are asked to mimic someone who is cis-gender and white and North America and European, then that is the femininity and masculinity that you learn, or voice-wise right. It's, it's white, um, European in femininity and masculinity, and that is also reinforced in the imagery. So all the there's a lot of like pictures in these apps and a lot of colors, stereotypically like red and blue, pink and blue. Um, and the people in the app, there are a lot of stock photos of like women in dresses and men and suits and, um, like mouths, like lipsticks mouths. It's like a common visual motif and these apps and there's like only like three. And so two out of, three of them have lipsticks mouths. 
+
+Alex Ahmed (00:12:37) - So I'm saying it's common, which is weird, right. Like why, you know, is it because voice equals mouth plus woman equals lipstick? I mean, it's just, it's very uninspired. Yeah. So that's, I guess intervention, intervention C or whatever. The third component of the intervention is like, let's just clean up this design. Like let's make it not weird. And, um, you know, not sort of make it not like the user has to either identify with these images or look at them and be like, Oh, why is that there? Just from a design perspective, we wanted to improve on that. 
+
+Joel Stern (00:13:24) - I mean, I was, I was interested in, um, going further into, um, the way in which, um, into the functionality of, of project spectra and, and, you know, cause I had to look at the, um, get hub and, and just, I'm just not, I'm not sort of quite technical enough to sort of understand like the processing and the algorithms and the different sort of capacity of the, of the application and what it does. But I also wanted to, um, just go back to the, the methodology where you were talking about, you know, doing staging a lot of interviews, um, with trans people about, about voice. And I was just wondering if, um, you could say something about what, you know, what you learned from those interviews and ha and how they, um, helped kind of set up the objectives of the project. And then, you know, perhaps that could sort of lead into some of the functionality of the application itself. 
+
+Alex Ahmed (00:14:19) - Yeah. Happy to. So, um, I did 10 interviews with trans people and code for the Boston area, and these were about an hour long and mostly in person. And, um, I asked them about just their general, like history with the voice, like have they thought about it? Have they used any, have they been to a speech therapist? Have they used any technologies? And if they have, I just probed to understand what that was like for them, you know, just sort of at a basic level, like emotionally. And so, uh, from that sort of going into, um, uh, sort of like speculative, like design, like.
+
+Alex Ahmed (00:15:09) - Question by saying, like, if there were something that you could use, like what would it do? And so while a focus of it was definitely like, let's figure out how to design a thing. I was really, or mainly interested in the emotional or like the like life experience of these folks and like where they were coming from. And, um, there was a lot of variety in people's responses. Some of whom were kind of on the fence about whether or not they wanted to pursue voice. Um, some were kind of just angry about it. Like I like don't like that I have to do this. I feel like it's something I have to do rather than something I want to do. Others are, we're really invested in it on a personal level. And some of them were, um, sort of midway. Like they were very, they were like skeptical, but they were also like, there are some things here that I want, and there are some things that I'm rejecting. 
+
+Alex Ahmed (00:16:15) - So for example, those one trans woman who said that, um, there are some aspects of voice that are sex and some are gender. So for her sex components of the voice are like things that can be related to like biology, like, like pitch and resonance are sort of like, you know, generally speaking, like cisgender women have like, you know, shorter vocal tracks or whichever. I actually don't remember exactly what the anatomical relationship is, but for her, like these were things that were squarely like sex. On the other hand, she said like other things are gender, like uptalk is gender vocal fry. So like talking in a sort of gravelly low, like, like a voice that's also gender. Um, and so she, she wanted to focus on what she called the sex aspect. Um, not everyone certainly is going to agree with that distinction or separation. 
+
+Alex Ahmed (00:17:22) - Um, but it's, it was her way of thinking about it and her way of deciding like what to do and what not to do. I also found that people were sort of creatively using existing technologies to like jury rig them into things that would work for them. So since there were really no like apps that fulfilled their purposes, they would just use regular like recording tools or they would use like apps that, that they could sort of just have on and just look at throughout the day to sort of see where they are in terms of their voice. And these would just be like general purpose, like pitch measuring tools. And these same tools would be things that their speech therapists would recommend because there were no like other options. So we immediately saw like that as like, Oh, like it's clear that even speech therapists are just suggesting people use pitch tuners. 
+
+Alex Ahmed (00:18:25) - So there is a, there is a way for us to like to move here that and do something that's useful. So that became kind of, um, one of our goals, which is to like, have something that didn't necessarily like give you an entire like lesson plan for like how to train your voice, but rather just like a companion that like someone could use if they wanted to, to get like a, a sense of where their voices and also to, um, to strengthen their voice, regardless of whether they're like going for. So we ended up with a set of voice strengthening exercises, which included like holding a note and that note itself, isn't like, there's no like gender component to that. Like you really just choose whichever one is the most comfortable for you. And it doesn't matter actually, which one you choose. It's just an exercise that is meant to strengthen like your, your vocal chords. 
+
+Alex Ahmed (00:19:28) - So it's not unlike what a, like a seed student would do. And so that, that was kind of intentional. Like at first glance, it's sort of like you open it up and there's like, you know, it just says, like pick a note to hold. And, uh, we did that because, um, a lot of the existing trans voice resources, both made by clinic clinicians and those made by trans people like, you know, this is the community involve a lot of repetition, a lot of exercise, a lot of training. And we were sort of in a position that we didn't want to endorse or denounced like any existing, like.
+
+Alex Ahmed (00:20:09) - Resource, but instead we were just like, here's something that like, could be improved, could improve your vocal health. So that's where that's the direction we decided to go. And part of it was because in the interviews we saw like so much variability and like what people wanted, that there was really no way to make something without, um, sort of having a, a stance on like, Oh, we endorse this. We think that like this path towards like vocal femininity is like better than this other one, because also like we wanted users of really any or no gender identity to like, feel like they could use it. Um, which is not something that, um, either Listening apps allowed for. It was most, it was pretty much just either you want a male voice or a female voice quote unquote, and you can see that in, um, in the apps that exist now, which don't even like express the possibility of being non binary at all. So, yeah. 
+
+Sean Dockray (00:21:17) - Side of the, like sort of tool you're hinting at it already right now, but outside of the tools and sort of exercises, I'm wondering how project spectrum as an app sort of does some of this like pedagogical work or how it sort of participates in sort of ideology, ideological kind of construction, you know, in the same way that you have been criticizing that the other apps have, you know, the sort of Midwestern SIS, you know, women as personality like, um, and the lips and everything that obviously there's a, there's a huge component of sort of like design as to how to structure how those exercises are laid out or how the kind of narrated within kind of this overall program, you know, of your own making. Um, I'm just wondering, like what, what, how did those discussions go with project spectra and how did you come to the, how do you come to, um, agreements or about what the sort of like graphic layout and what the kind of pedagogical kind of dimension of the app and how to try to find your way through these exercises is how does that work? 
+
+Alex Ahmed (00:22:29) - Yeah, that's a good question. Yeah. So, I mean, just like trans people, like in general have a lot of different ideas about like what boys should intake, like what voice training should or should not entail. Like, so to like our group was really heterogeneous and there definitely were like moments of disagreement. And, um, I think that played out in a lot of, um, complex ways. I think for one we sort of wanted to do away with pretty much any, um, symbology of, of gender. So like we don't have any of that. It's pretty much just, it's a clean layout. Like, you know, non-gendered color schemes. Like it's like mostly like greens and purples and oranges. I'm a fan of that color scheme, but anyways, and, uh, the, the language is very sort of supportive. Like that was one of our goals. Like we wanted users to feel supported, but not judged or like condescended to, so that kind of met, like I met a few things like one, um, you should be able to sort of be in dialogue with the app rather than it being like, this is what you should do. 
+
+Alex Ahmed (00:23:54) - Like you, you need to be at two 20 Hertz and you need to repeat after me. And, you know, so instead we kind of have it, like the beginning of the app is like a series of questions. And so, so the beginning is like, you know, what's your name? And then, you know, you put in your name and then it asks you, what do you want to do with your voice? And so you are, you're allowed to pick from a series of options. So that's, I'd like my voice to be more androgynous. I like my voice to be more masculine. And I like to be my voice to be more feminine. And I don't know, I'm just exploring. So you're not really like immediately like shoehorned into this like trajectory. You're more like, um, we wanted it to be like sort of a, a branching path where like exploration is like encouraged and like play and like expression in a like sort of fun way or encouraged. Um, and the degree to which we actually accomplish that, like, you know, can definitely be up for debate and, you know, a continuous process of revising and iterating, but that those were our intentions.
+
+Alex Ahmed (00:25:09) - To, um, to have that be the mood. And so users can, um, pick, like, if they have like a specific goal for where they want to be, like, does that be two 20 Hertz? It could be anything, you know, in any range, like, and then like throughout the app, like your, your pitch as it's measured by app is like, is going to be like, compared to that, um, that, that self selected marker. And then alongside that, we just have these completely gender agnostic, like strengthening exercises, um, which we suggest within the app that you do, like twice a day. Um, just so that like you're not straining herself. So any like suggestions that we make are purely from that lens, just, uh, you know, don't hurt yourself. It's okay. Like take it easy. Um, so that's kind of like the, the tone of it is just like, um, you know, make sure to take breaks, take care of yourself, like, remember that pitches and everything. 
+
+Alex Ahmed (00:26:18) - And this is just, uh, you know, a way to, to reduce strain. That's sort of what we ended up with. That's the answer. The other part of your question there, there definitely were folks on the team who one of their, to be a, a complete guide, like a sort of step-by-step like, this is how you feminize your voice. And that, that was their, um, their interest as a trans feminine person. So that's the guide that they were working on a complete set of exercises, a Hutton entire pedagogical roadmap to like get to like a, you know, a feminine voice. So it was there, it was their wish to, to create an app like that. And part of that included, um, exercises that, um, that we didn't end up including in the app, but that like we had created an algorithm to, um, to detect. So that's a residence exercise, which the trans voice training, internet, community calls, big dog, small dog, where you pant like a big dog and then pant like a small dog. So that looked like you could try to do, you know, if you want it's like, so the idea is when you pant like a small dog, you're, you're like sort of configuring your like vocal resonance in a way that like primes it to like, be more feminine because the, the vocal cords and the musculature of your mouth is shifted up in that way, because you're, um, if you can like, feel your larynx, it's like, if you like, kind of like a small dog, kinda like a small dog, 
+
+Alex Ahmed (00:28:17) - You can feel it going down. So the idea is that a raised larynx is good because that makes your resonance, um, more, um, like more bright, is it like the, the speech science term for it? Or if you paint like a big dog, then your voice is like, your vocal is darker to use that, to use their phrase again. So some larynx, X practice and larynx training is like very common in these online communities, which project spectra as an online also, um, we're, we're like drawing a lot of these same people. And I can get to that later as sort of like a limitation of our process, because those are the people we worked with. Those are the people we found. So the larynx exercise is like very disagree. It's very controversial in like the speech therapy world where some speech therapists would, would say like, don't do it ever. 
+
+Alex Ahmed (00:29:22) - Cause actually it could hurt your voice. And some might say like, Oh, do it. But like only if you like really want to, or like, if you are, your specific goals would be student to it and would work with you to like, get to, uh, get to an understanding on that. But, but otherwise like online, big dogs, small dialogue, or like meringues exercises are, um, like pretty much universally recommended. It's like, do this. If you want to sound feminine, do this, do you want to sound masculine? Like you need to, you need to have control of your lyrics in order to maximize your potential. And, and like, you also see this in like singing, singing teachers will say some more things. And there's also similar controversies about, about lyrics, like movement and training, um, in that world. So.
+
+Alex Ahmed (00:30:10) - Basically, um, I, to get, to make it more concrete, I was working, um, with a speech therapist in the Boston area during this project at the same time as I was working with, um, these folks, the project Spector folks online. And so there was this disagreement where like, I sort of came to my collaborator and I was like, Hey, like, you know, the folks I'm working with, they want to do a larynx exercise, um, include that in the app. And she was like, Oh, I wouldn't recommend that. And so I was like, Oh, okay, this person's on my dissertation committee. I kind of like to make sure that she's happy the same time. I don't want to like, sort of betray my ideals of like, wanting to make this app like designed to like buy in, in the community. Um, and so like, that kind of brings me to like one of the central contradictions of like this work. 
+
+Alex Ahmed (00:31:09) - And, um, and that was like the fact that I was very much like in the academic world, like being, um, sort of held to all this academic, like rules and standards and codes and like, you know, deadlines and deliverables that I needed to do. And so a lot of the suggestions that came from the community that I was working with actually didn't end up in the app that wasn't true universally, like definitely the app like was designed by and coded by us, but like I had a hand in all of it, you know, so to say that, um, stay that it was a sort of pure process would definitely be, um, would be false and I definitely can't claim it, so I won't, but yeah, so it came through in that way and that sort of, you know, the debates and discussions we had, um, kind of led us to be like, okay, well, you know, we'll leave it out of this version, but you know, maybe in a future version you can put it in, you know, knowing that like, you know, because it's open source, like we could just keep adding to it, like, um, beyond the life of the projects. 
+
+Alex Ahmed (00:32:23) - So, so that's kind of where we ended up 
+
+James Parker (00:32:25) - And that's a nice segue actually into the next question I was going to ask, which is, you know, how, how has it been taken up and what futures do you imagine for it in terms of scale or design or anything? Really. 
+Alex Ahmed (00:32:39) - So when, when I came out with the sort of first release, which was in October of last year, it was kind of like the beta and we were going to like share it with people, get their feedback. And the same time I, as a requirement of my degree, I had to do, um, an evaluation study where I had to give the app to some people and get their feedback and opinion on it. So I kind of thought, okay, like that's kind of a reasonable, like, like way for this to segue into the next chapter of the study. And in fact, like the app had to be done by that time in order to have enough time to run a study. And like the few months I had left before graduation project spectra kind of like stopped being involved in the evaluation study. Cause that was like, Holy done. 
+
+Alex Ahmed (00:33:37) - Like in Boston, like through people I could recruit in person, um, and like get approved through like the institutions, like review board and stuff. So that formal part of the study happened and was happening right. When hope it hit. So my dissertation kind of ends early and that like, I, um, I had had two people who use the app for a month and I interviewed them after that month. And I started got their sense of like how it worked for them or didn't work for them. And so that was kind of the ending of the project. Like I wrote it up and defended my thesis and then I sort of moved on to the next chapter of my life. And there was some movement in project spectra even after I graduated. So there was like a company, a startup that was like interested in like working with us and who like was talking with me. And I was sort of like saying like, don't talk to me, talk to the group, like, you know, I don't want to be like a leader of this. And so I kind of was encouraging them to like, Oh, like just talk to the discord, like community and see if they want to help you, like, you know, flush out these ideas more because I kind of felt that I had sort of lived out in my like stint as like the de facto, like.
+
+Alex Ahmed (00:34:59) - Person calling the shots, like person saying like, actually this is what needs to be in the app. Cause like, Oh, I got it. I got to finish it so I can like defend and stuff. So I sort of felt like I wanted to like, let go a little bit with the hopes that like, Oh, like maybe like things will get taken up, but it kind of didn't unfortunately. So, um, I mean, maybe COVID has to do with this, but like, you know, after March or so, like when I finished my dissertation study mostly and uh, graduated like a month or two later after I had written pieces, it's been pretty quiet on the discord, which is a shame and I'm not really sure what to do about it. Um, because I would like it to be seen by more people and to, you know, to get more feedback on it. 
+
+Alex Ahmed (00:35:50) - I know that like, it's definitely not done, you know, it does a few things pretty well. Um, but like the people who I interviewed who had used it for a month, um, they definitely had a lot of criticism and feedback and a wasn't universally negative. Like they did like some things about it, but yeah, like to claim that it's like a finished product, it's definitely not, it's not, definitely not true. So I, yeah, I mean, I would hope to keep doing it. And part of why I had stopped was because I needed to get another job because I had finished my degree. I didn't have any, every job, I didn't have any income. So I started doing part-time web development, like a contract gig. And I felt like, cause I had not on the academic job market too. And um, I had struck out I didn't get anything. And so, um, I was like, you know what, fuck it. I'm just gonna like go into industry. I'm gonna like, just do whatever my boss tells me in KB. 
+
+Alex Ahmed (00:36:53) - And, um, and yeah, like sort of leave this sort of academic life behind and sort of just like work on whatever projects I want. So that's kind of where it's at. I mean, I, I think I might like to work on it again someday, but I guess I kind of would just want to hope that like the community will sort of revive again, cause there was a lot of people involved in it and on the channel, I think it's just that like, without like a direction, it's just people sitting in a channel and since, since it's discord people, aren't inside the channels doing other stuff and like talking to other people or doing whatever they do. So, you know, there was not really a impetus to do anything right now. 
+
+Sean Dockray (00:37:37) - How's he going to say? I think that's such a common, not common, but I mean, it is a sort of a common, like a open source project situation. I found myself in it personally, a number of times of, um, you know, you build these community projects and try not to be sort of, um, a leader let's say, or a spokesperson or, you know, like you want it to have a momentum of its own, but clearly you, you invest inject some momentum into it. And then, um, it's not so easy to pull yourself out and not have all that momentum sort of disappear, but I am interested in this next step that you've taken. I mean, I realize it's, um, it's a little bit just a, um, a consequence of the academic job market being just like horrific at the moment and just in general, but also, you know, at this particular moment, uh, and that you find yourself in kind of web development. 
+
+Sean Dockray (00:38:29) - Um, because like I actually think it's interesting cause I, I also sort of need on occasion to, um, like actually solidify my, my income situation through similar means. So I, I empathize and recently that kind of world has been taken over by this like a user centered design IDEO kind of, uh, rhetoric, which you talk about in the methodology of the paper on project spectra. And, um, what I'm trying to do is connect that to something else that I've found, which is recently in those worlds, you know, because, um, voice interfaces are kind of taking off just as a, as a, you know, there were, there was like websites and there were apps. And now I think people are like a lot of these web firms are kind of developing for voice specifically. And so you bring a really interesting and important perspective to that. And I guess I was just wondering, you know, you can pick any part of that question to talk about, but what I'm interested in is either you expanding a little bit more on this kind of like critique of user-centered design, as well as, um, thinking about as it kind of applies itself to voice interfaces and Machine Listening more generally, like how, how you sort of see that playing out, particularly with your experience of doing projects, spectrum. 
+
+Alex Ahmed (00:39:52) - Yeah. So actually, um, today was in a meeting with some grad students.
+
+Alex Ahmed (00:40:01) - And, um, for context, I'm a postdoc now at Carnegie Mellon and just started, I don't know how I got here, but I'm doing that now. Um, but so I was in this reading group of, for like, um, tech injustice, which my advisor, Sarah Fox runs. And so one of, one of the, um, things we were talking about in that, um, in that reading group was like, what is the responsibility of like an individual designer? If the technology is doing something socially deleterious. So like, say for example, um, you're creating like a machine learning system that is going to end up determining, like, who goes to jail longer or like who gets healthcare or like whose insurance claims get denied or whatever. And so I think that the, um, like the user centered design IDEO model kind of like keeps us in this world of like the individual user and like, and like the individual developer who has these tech skills, who's like trying to like understand the user and like, you know, who is the user? 
+
+Alex Ahmed (00:41:18) - And like, it, it, it sort of like constraints the, the world of inquiry into like, just these two parties. So for example, like you end up unable to talk about how, like, you know, like user centered design is becoming like prominent in like the military. Like, and so we ended up being like absurd and saying like, okay, how do we like maximize user centered design principles for like war planes? And it's like, um, like who is the, okay. The, yeah, like, okay, making life easy for the user. I E the people flying the fucking bombers, like is not going to be good, like that's actually bad. Um, and so like, and the, and like, I think like students, like, I kind of like want to approach, like, I, new role is like slightly higher on the academic totem. Like, I shouldn't be set up on the academic ladder to like, be like, you know, let's like nudge folks, like, sort of thinking more like on the systems level. 
+
+Alex Ahmed (00:42:28) - Like, and if, even if we're thinking about politics, 
+
+Alex Ahmed (00:42:30) - Like maybe thinking about politics and like, Oh, not Leslie neoliberal way, but it's hard because like the dominant ideology, the field of design user centered design is user centered. And it's not like really talking about like these problems. And so you end up 
+
+Alex Ahmed (00:42:48) - Stuck in like, Oh, like, what am I as an India, 
+
+Alex Ahmed (00:42:50) - The visual designer they're going to do to like, make this better. And that can be like really, um, problematic, I mean, in, in more than, than what I just mentioned. So, yeah. I mean, I think that, like my wish my, my like, sincere hope would be that like, people think about like these things in terms of like, not like, how do I do a better job, but like, who controls what? And like, who has power to do what? And like, if we think about it in those terms, then we can think about like, okay, actually we need to, we need to be fighting against this. We need to be like, fighting against that. We need to organize in this way. We need to strategize and like make collective demands of these people. So, and, but that's like, that's organizing talk. That's like union talk. That's like, not really something that like, people want to hear, especially these corporations that are shelling out big bucks for like, you know, design experts now. So, yeah. I mean, I feel like the corporatized, like design world is definitely like intertwined with the academic design world. And I don't really know, like, if there's any real solution to this aside from like just burning it all down, but here we are. So, yeah, 
+
+James Parker (00:44:17) - That's a great answer. I mean, it's, so it's so sort of simpatico, I suppose, with what we're, who we're trying to, you know, the politics of this project. Do you have any thoughts on the sort of the specific voice user interface or Machine Listening angle, um, or even opening out to think about those sort of agenda politics of voice assistance or Machine Listening more generally? 
+
+Alex Ahmed (00:44:41) - Yeah. Um, this is weird stuff. So like I recently saw a, um, Twitter thing that was like a new, like gender lists, uh, voice assistant.
+
+Alex Ahmed (00:44:55) - And they listened to it. And I was like, uh, I mean, it's like, kind of sounds like meaning, like, and so it was like this weird feeling where I was like, I'm not a generalist. Like what, so why, so I feel like it can like this sort of drive to like, degender, everything is like, really, I kind of bothers me because like, I mean, not to say that, like, you know, we shouldn't like destroy this fucked up gender binary system, but like also we can't just like, pretend it doesn't exist, you know? And that, like, you know, if I'm listening to something marketed as gender neutral and I, and I have like opinions about it that, you know, I, that I'm gendering it in my mind, then it's like, what, what have we really accomplished? I mean, like the, the feeling of gendering something, or like being gendered is like not affected by the like, desire to get rid of gender. 
+
+Alex Ahmed (00:45:57) - And like, I feel like that can affect, like, I don't know, folks who, folks who do like, feel that this is again, misrepresenting them to see like, Oh, like this isn't gendered voice. And me thinking like, well, is it because I know a lot of women who have voices in that register, I know some men who have voices in that register, like non-binary folks would also, I think I can't speak obviously for them, but just like knowing, knowing a lot of non binary folks would say like non binary is not the like mathematical, some of like male and female, or like the mathematical, like overlap of male and female, you know what I mean? And so 
+
+James Parker (00:46:46) - Is it the genderless voice assistance or a voice user interfaces sort of buy into an ideology of pitch? Basically they sort of while like trying to opt out of certain kinds of ideologies agenda that they've actually re inscribing the one that tethers gender to pitch. 
+
+Alex Ahmed (00:47:07) - Yeah. And like, and I mean, like, I, I'm not like saying that I'm above this, like our app also includes this because we wanted it to be useful in that way. I mean, even if folks, if folks were interested, they have the ability to like, not participate in that. Um, but like if, if they did want to and couldn't, and that would also be a design issue. So the idea that you can like opt out of ideology, I think is totally wrong, you know, I don't think it's like not a worthwhile like, goal to try to do like something that is better, but, um, I think it would have to be something more than let's pretend it doesn't exist because it does. I mean, that's why I 
+
+Joel Stern (00:47:54) - Think, um, when you were talking before about sort of exploration and experimentation, um, as a core value of, of spectra, it's in a way that works against the sort of normative, you know, inscription of say pitch to gender, because it, it allows you to arrive somewhere that is expressive and perhaps, um, also liable to change according to however you feel is sort of from day to day or, or et cetera. And I was, I was sort of thinking before also about, you know, the relationship to singing apps and the way in which, you know, a great thing is throughout history have often circumvented gender expectations or, or had totally non-conforming voices, which then sort of signify a certain kind of imagination and creativity. And so it sort of seems like that's part of, um, if not an answer part of the, you know, methodologically, something that, that we should be valuing here. 
+
+Alex Ahmed (00:49:06) - Yeah. And I think it kind of like brings back the focus away from like, what does, what do other people think of me? And like, what do I want to think about myself? Which is something that like, like myself, I've dealt with a lot, like, you know, just like understanding like, okay, how do I fit in, in all of this? And like, to be hearing so much of like, actually you're not trans because XYZ, or like actually, like in order to be trans to have this XYZ, like childhood memory or like, you know what I mean? So like, instead of that, or like, you know, even in the context of boys, like in order to have like a feminine voice, you could have XYZ. So instead of that, like to say, like,
+
+Alex Ahmed (00:49:51) - How do you want to express yourself? You know, how, how do you want to, like, how do you want to be like, um, in, in your life, you know, like how do you want to, um, to present? I think that doesn't sort of say like that these norms don't exist instead. It's saying like, we know the exist, but like fuck them. Right. Like 
+
+Joel Stern (00:50:18) - I think, fuck, it might be a nice, a nice note to end on as well. If we cut the recording right there.

content/transcript/billyli.md

@@ -0,0 +1,80 @@
+---
+title: "Jùnchéng Billy Lì"
+status: "Auto-transcribed by reduct.video"
+---
+
+James Parker (00:00:00) - Thanks so much for speaking to us, Billy. I mean, would you, would you maybe introduce yourself and just say a little bit about the kind of work you do, how you ended up doing that work, where you do it and so on? 
+
+Jùnchéng Billy Lì (00:00:10) - Absolutely. I am currently a PhD student and Carnegie Mellon, computer science departments, language and technology Institute. So our departments, uh, like a main focus on research Lising NLP and then speech and audio analysis. So, which also is what I focus on. Um, you know, as the, my major research goal, which is about, um, audio and machine learning, basically. So, uh, previously for a year, from, I think a year ago, I, from 2015, til 2019, I was a research engineer working for Bosch research technology center in their AI department. So I was, that was a joint collaboration between Bosch and, uh, CMU. So I, I worked for them in a very similar domain where I worked on exactly Machine Listening and sound event recognition for the use cases Bosch had for their industrialization of say Machine health monitoring, and also, um, smart, like a smart ear project, which they define us to launch potentially a Bosch device that can listen, uh, L listen for ambient sonnet event. 
+
+Jùnchéng Billy Lì (00:01:35) - I don't know what exactly that is going, but I left Bosch in 2019, but the adversarial robustness aspect started in 2019 where we were generally interested in exploring whether, as I said, you know, the current machine learning techniques for sound event recognition and for speech, um, automatic speech recognition is actually robust in a real world setting where we are free from, you know, the, uh, the influence generated by ambient noise. That's ambient noise is what we cared about in our work to really, um, try to answer sort of question. Can we demonstrate that the current mesh and machine learning technology or the machine learning models, um, are they actually robust against these ambient noise in a real room setting where we can demonstrate a specific case that they can break down in certain scenarios or if the best they are actually robust to any type of a part of patients. 
+
+Jùnchéng Billy Lì (00:02:45) - But what we found is that, you know, was a certain decibel of sound, which is the artifact that we generated using the techniques that we presented in a paper, which is basically the projective grade in descent methods. So we can actually replay this piece of audio and then trick the machine learning classifier to think that the, uh, so the way court in our case does not exist in like in a real room setting. So to kind of like in a word, basically it shows that the current state of art machine learning technology or machine learning model that we deploy in a, in a smart speaker or smart devices are not robust against a certain type of a curated noise that is aimed on breaking these models. That's well, what our research show as an early stage, finding how to improve it, it's still an open question. 
+
+Jùnchéng Billy Lì (00:03:48) - We can train a model to lead these devices. Remember these are the adversarial or the audios, or these are the potential noises that breaks us so that it doesn't get trick the game, but it certainly doesn't guarantee these models are robust against other type of noises where we can change the threat model. So basically as an attacker, you have an afford information of a model, or you have a, you are more flexible than the defenders. The defenders can only defend against known attacks, which are actually very predictable, but as an attacker, you have a freedom to do many different things where the defender are not necessarily, you know, able to foresee this, and then they are less flexible in terms of defending it where it comes to like specific security, uh, sensitive applications. It could be a real trouble for these type of applications. 
+
+Sean Dockray (00:04:48) - Could you talk a little bit more about what some of those threats might be? Just, um, just to give some context as to why people might be testing. I mean, why people might be attacking, um, some of these devices or some of these machine learning models.
+
+Jùnchéng Billy Lì (00:05:07) - Yes. So what we like the motivation of our work, or because as of now, now I can speak as a total, like a CMU student because I left Bosch. So basically from a research standpoint, the motivation comes pretty strong. And then, so as we can observe in the general tech world or tech society, basically Google, Amazon, Apple, they all have voice assistant stored was a bunch of your own credentials, which is a huge, like a liability of, uh, your kind of a personal information stored under sound enabled devices. It means specifically, for example, Amazon, you can order your goods from Amazon. Alexa was only two piece of audio, um, commands basically say, Alexa, can you order this and put on it, put it in a car. And these things will directly be linked with your credit card info or your personal account. If you're you got a annoying child, for example, at home, basically they will order like a bunch of random stuff and you don't expect these things to happen. 
+
+Jùnchéng Billy Lì (00:06:24) - But so here's like a, these are the extreme cases, but in general, like, uh, audio monitoring is happening more and more. If you guys have looked into it, like basically we have these monitors, like almost everywhere. It's kind of ubiquitous with cell phones was, uh, was, uh, was smart. Speakers was nest. Webcams was, uh, home security, doorbells. All these things have sound recording. If you guys have looked into it. And then these certainly created a huge, like a research Corpus for people like me, where we can collect huge amount of a personal audio sources where we can train our, you know, fancy models on to tell whether these are, you know, events of interest. So to say, because like, for example, for the audio event, detection is like, if you have environmental recording from a doorbell for like a month or say three months, and you can basically tell what are the ambient sound going to be like, how is it going to sound when someone knock on the door or approach the approach, the, uh, you know, the garage or, or basically walk down the floor, like all these things have its own signature. 
+
+Jùnchéng Billy Lì (00:07:45) - And of course everyone's speech has his own signature. If you're a model can pick up like a recording of somebody long enough, you can train a model to totally mimic that person was speech synthesis. So the general rule of some right now is like, if you have a hundred hour clean data of someone, I can pretty much use taco troll or like Wavenet to regenerate your son and then basically fake your voice for a phone call. And you will not be able to tell I am actually a fake person. Whereas, you know, as long as the video doesn't show up or you don't get other modality of input to verify that you are actually yourself. So in my opinion, these things are very, very safety critical where people, the public, as you guys observed have not actually looked into this because audio has not yet become like a such hype as autonomous driving because you know, these things are less conspicuous, but is actually like scary in a way that is ubiquitous because the recordings are happening. 
+
+Jùnchéng Billy Lì (00:09:00) - They're all around us. And then, but we're not very aware of like, they are actually listening to us. Like for example, the us case they are, I think they have an algorithm like of recording constantly so that they can actually wait for the way court. The wake word detection model is actually Listening all the time. So I am not sure whether these data gets sent to the cloud of Alexa or Google home constantly, but I'm sure they definitely kind of opened a back door to record for a period of time to collect these datas. Otherwise they wouldn't have enough training data to train the wake word, which is very troubled for some, for their application. So from a, yeah, from a legal perspective, I'm not an expert, but.
+
+Jùnchéng Billy Lì (00:09:52) - I'm sure like, you know, whatever they claim, they need to collect training data for training, uh, to, to train their, their model to make, to improve their model. So there's gotta be some way for them to acquire these data, basically that's up to like, you know, to, to the public, to, to really look into their terms of the con and conditions about using their smart devices, which I haven't looked into it, but from a technical aspect, I'm pretty sure that they have to acquire at least thousands of hours of good quality audio recordings in order to train a model that is of a satisfactory performance. So, yeah, to sum up, I think like these things are very sensitive right now, which not many people actually have put their attention onto these sound related issues. So I'm really glad that you guys have caught these things. So, um, and then we can talk about these things if you guys are interested. 
+
+Joel Stern (00:10:55) - Yeah, absolutely. I mean, I think you've, um, um, covered many of the things that we're specifically interested in, so, and it's, it's great. And it's also, um, great, great to hear them from, um, the perspective of, of, of a research, uh, you know, we're working on the technical side of, um, these questions just, just wanted to, but I think we want to move into sort of talking about more specifically about adversarial or, or do you think in a minute, but I just wanted to ask you when, when you were saying, um, with Wavenet and sort of other applications, you know, there's the potential to reproduce a human voice in a way that, um, is indistinguishable from, from that person's voice, is that, is that indistinguishable to, to a human or indistinguishable to, to sort of other Machine listeners or, uh, I wonder if that there, sort of that question sort of continually comes up about the difference between the way in which these sounds, uh, kind of, um, heard by humans as opposed to the, to, to the Machine listeners. 
+
+Jùnchéng Billy Lì (00:12:00) - So I also can tell you my experience of a Wavenet. So if you have like 100 an hour of a recording of you say, like, you know, in your daily life, then I can pretty much guarantee that for a simple sentence. I don't think you would be able to tell, like from a wave next generator sun, then like your real human voice, there's only a very subtle differences between these twos because Wavenet is, uh, driven by the same techniques as, uh, audio events or the specific audio events where they break down each element of recording or the generation into phones phonemes specifically. So basically whatever you pronounce each full names were in Alexa case, it has six, four means of, uh, a cup saw. So it's like all these six phonemes get picked up differently. And then the machine learning model is able to tell or regenerate in wave that's case like regenerate, how exactly you pronounce these things. 
+
+Jùnchéng Billy Lì (00:13:20) - It might sound mechanical if you don't have enough training data to train these things, but given enough data, basically it can sort of, uh, compute a language model. In our case, basically language model stands for a probabilistic model of predicting what is the next full name in the prediction. So yeah, it captures the likelihood of the next outer ends of your sound. So it's kind of intuitive for the public to understand that, you know, basically if you capture long enough, your daily words like, Oh, daily active vocabulary, then it will be able to regenerate or whatever you say, things in a certain way, as long as you don't exert a very distinguishable emotions or like was a exaggerated way of speaking. So it's kind of a, yeah, it's kind of dangerous that, you know, if someone get access to hundreds of hundreds of hours of speaking data of yourself to store it on their cloud, they can actually, you know, if people really do have a, a malicious intent, then they can really replicate these. And, uh, I think if you guys are interested, I can point you guys to like the Wavenet demonstrations that have their official website. And also some of the experiments that my colleague and I have run. So you can try to tell if.
+
+Jùnchéng Billy Lì (00:14:55) - They, if you guys can actually tell the difference between the human beings and then the Wavenet generated, 
+
+James Parker (00:15:01) - That would be great. I mean, I've spent some time with the Wavenet, uh, the, you know, the official website, um, but you never had know exactly how much that's curated, you know, like that, the famous example where they demonstrate, you know, the, uh, the way Google demonstrates, you know, booking a haircut live. Yeah. Like I just, you know, obviously that's sort of amazing in the video, but you never quite know how, how curated that is. Right. So, I mean, it would be fascinating to listen to sort of an independent researchers, uh, you know, attempt. So 
+
+Jùnchéng Billy Lì (00:15:36) - I think I can definitely look into several related links that you guys can potentially play around with and then see the real effect of a Wavenet. And then the taco troll, if you guys are familiar, was that paper. So it's basically the techniques behind speech synthesis at Google. So, and then many people are doing similar things to regenerate, like a piece of audio in that way. So it really like from the role, like, uh, audio feature perspective, they're really captures the role raw features of your sound and a regenerate bunch of the similar features of, uh, your speech and the, basically the source feature gets perfect match was the target generator features. So you can't really tell these things are actually different because humans, son, like a human sound perception is built around the cochlear nerves, which the sound recognition systems are built to mimic those things as a filter banks where they mimicked like were more sensitive to like low frequency sound, but less sensitive for the high frequency sound. So these features get perfectly captured by, by the synthesizer where you can actually kind of get tricked to classify, like if this is Machine generator of sound or, or human generated sound, 
+
+Joel Stern (00:17:10) - I'm glad you, you mentioned, um, trickery, uh, because, uh, I think it would be good if you could sort of give a kind of just a general introduction to adversarial audio, you know, what it, what it means, um, you know, what, what it sort of does, how, how it works and, you know, because, uh, when I, when I look at the demonstrations and the examples online, it, it sort of is a form of trickery. It's, it's sort of tricking the Machine it's, it's, it's tricking the human, but of course underlying that trickery is sort of complex computation and modeling, but w w would you maybe be able to sort of, um, give a broad Def definition of what, what adversarial 
+
+Jùnchéng Billy Lì (00:17:54) - Audio is and how it works? Absolutely. So basically, so in a nutshell like adversarial audio by means is basically trying to be adversarial against a machine learning model rather than human beings. So we are targeted to trick a machine learning model, which is deployed on smart speakers or a piece of a computer software that is able to recognize certain sounds as they claim. So, for example, in our case is Alexa echo devices, where they claim to be able to listen to the keywords, Alexa all the time and, uh, respond to your human comment. So that model is our target of interest for attack. So the way we attack them is basically we try to understand how that model works first. So we kind of looked into the publication of a, the Alexa team where they, they, they had revealed a certain technical parameters of their models over the years where they exposed to the public, how they train the model, what kind of dataset did they train on? 
+
+Jùnchéng Billy Lì (00:19:13) - And then what kind of techniques, or the specifically the architecture of the neural network models, they have used to train these models to tell, like whether there exists in a wake word of Alexa in a, in a ambient environment or not. So that's the model that we're attacking. So the first thing we do is emulate the model. So we build a very similar model using the public available, um, pipelines for machine learnings. And we collected 2000 samples of a wake word that through our own kind of data acquisition pipeline, where we kind of went out to collect these data as ourself, where we have a bunch of friends record Alexa, basically. So we have like about 5,000, um, active cases. And, uh, we synthesize the rest of, uh, inactive cases for the ambient environment classes so that it will fire as a negative case. So this is the model that we emulated for, uh, for a fake Alexa or our own version of Alexa. Basically we.
+
+Jùnchéng Billy Lì (00:20:23) - We, we don't want to use the word reverse engineer, but it's, we actually reverse it engineer like a Alexa to some degree. But the key thing we want this fake model or emulated model to do is to really showcase the gradient or the technical parameters of the real laksa model. We expect these gradients of our emulated model will look very similar to the real Alexa models because the behavior of the models are the same where they fire as a true. And, uh, when, uh, when there is a wake word of Alexa and they fire as false when there is no Alexa. So then we can attack our own emulator model as a white box attack instead of totally running blinded as a black box. So tack where we don't really get to observe what the models are doing halfway. So basically assuming that we have a halfway or a generally good enough white box models for an Amazon Alexa, then we can really look into its propagation of gradient. 
+
+Jùnchéng Billy Lì (00:21:34) - So a variable, uh, uh, currently like a state of art machine learning models that are pretty much all. If it's a deep neural network model, they are all trained, uh, by the same tech sneaks of, uh, back propagations, which means that propagating the arrows of each layers of neural net to the latest two, to the surface layer, to, to recollect, uh, to basically to collect all the sum of arrow that you get from one data sample, where you can train, uh, a target, you can have us a target signal of a NeuroNet to let it converge to the, to the target that you want through the proper loss function that you define. So what we do is we don't want this loss function to actually decrease because normally when we train on you in an hour, we want the loss function to decrease because we define target. 
+
+Jùnchéng Billy Lì (00:22:36) - We want the model to converge to the minimum loss, where we have the minimum arrow, so that we will have minimum ever arrow in this case, detecting whether the true Alexa word is, uh, is it actually, um, existing or not? So when, whenever there is an Alexa in a ambient sound, then we will fire because our loss is very low. And then our model is actually having very little arrow of the, uh, of a finding whether this is actually true or not. But in our case, we're generating the adversarial audio where it act as an Delta to the original input X. So we kind of like add a piece of noise from, you know, Our own generated data. And what we're training is actually Training this adversarial Delta as, uh, the same shape or the same type of a feature was the original sun. And I'm confused the model to take in X plus Delta. So this Delta is our adversarial noise and then gets overlapped with the original sound. So this X plus Delta, what we're training it on is actually to, to maximize the original loss. So we don't want the original loss to actually go down decrease. We don't want their arrow to be small. We wanted to go the other direction so that you will amplify the arrow. So will is the existence with the presence of our Delta. Your model will not be able to decrease your arrow anymore. You will only go up because we have curated a bunch of data that we train on was many iterations to make sure that your original loss is not going down, whereas it's actually going up. 
+
+Jùnchéng Billy Lì (00:24:34) - So with our presence of, uh, was the presence of our trained adversarial audio. And next was your original data input. You are actually doing the opposite thing as your original model. So the, the arrow of your original model will not be small. It will be very big. So you get, you will be confused. And the model itself will be misclassifying bunch of the targets where it was supposed to do when it's clean data. So what we're doing is basically jamming the model to let him do wrong things because the X plus Delta is not the original acts where the model recognize or remembers to do the right thing.
+
+Jùnchéng Billy Lì (00:25:23) - We are confusing. The model to feed it with a wisdom was the presence of noise. It really gives the model that something they are not good at, or they will tell these are the wrong classes. So they will always misfire. So in Alexa's case, we feed the Alexa was an X plus Delta. So this X plus Delta is the, we cannot control the X because that's the ambient sun were in, but we can always control the Delta. So the Delta is our adversarial noise. So the L when Alexa listens from the MBN world, they're picking up X plus Delta necessarily because our Delta gets overlapped. It was the ambient sound environment that you are in, so that the X plus Delta is Alexa input. And it will get confused because we put our tricks in the Delta S term. So Alexa will misclassify whatever it was supposed to classify, which is the wake word in their case. So they will not be able to recognize the wake word is actually active when there is a way cord and plus our Delta, the way court has gone. So Alexa will be staying silent. The light will not be blinking anymore. So that's what kind of like a long story, but sorry about my, you know, verboseness, but I just want to make sure that the entire like logic guests, propagated, and then you guys are able to follow what I was saying. 
+
+Sean Dockray (00:26:54) - Thanks. That was, that was really thorough and really helpful to, to listen to just some quick follow-up or just to clarify, when you say the kind of ambient environment, um, I just want to picture the kind of process that you're going through. Are you, you're not sending the data directly into the model with the perturbations kind of on it, but that you're, you're talking about playing audio into a room, right? So the Alexa sitting in the room, so this is something that works in real space. Yes, 
+
+Jùnchéng Billy Lì (00:27:24) - Yes. So that is also one of the trick or why our paper gets accepted because these are the technical novelties or like say, I mean, the general idea is also was also novel so that it was, uh, actually a kind of early work that people had raised a bit of a community's awareness, because when you process audio, we are familiar with all these, um, room impulse transplants or, or response, or like transformations that a song gets process when it, uh, get recorded from the microphone. That goes to the same question. I think Joe asked me through email, whether it's zoom, we'll be able to do these things. I'm not really sure because every sound recording, advisor's doing a sort of a transformation of the original signal. So the rule of thumb is, uh, it always do FFT. We stand for fast-forward transform and breaks down your, your, your signal to kind of stratify your signals to different frequencies. 
+
+Jùnchéng Billy Lì (00:28:34) - So you di you, you can keep more information in the lower signals where there are human speeches, where you can kind of neglect the higher frequency, where there are more high-frequency noise. All of these, these things get taken care of in the algorithm space, where we, if we want it to in our world, we wanted to too, with, you know, we wanted to trick Alexa without really happy to it's hardware, so that we really have to do the same transform as what Alexa does. So basically we can encode this Delta, as I said, this Delta is already taken care of by the transform function that will be defined as the sound transformer. Uh, as a trustee can transform like a function. Basically it will handle the room distortion, and I will handle the echo distortion and also will handle the ambient angle distortion that you get from a real room generator. So when the digital microphone picks up the piece of, uh, audio that we generate, it's able to do the exact things we tell it to do. And then through the optimization process, we train on the network. We train on a computer. We are as if this Delta was generated in a physical space, so that the Delta, the noise that we train is already trained to wisdom, consideration of a being in a real physical space so that this Delta, when played out.
+
+Jùnchéng Billy Lì (00:30:20) - It's the real sound that it can real really play in, in the ambient environment where it doesn't get distorted anymore, because it already has get distorted in the training procedure where we thought that these things will happen when a sound played in an open environment in the, in my living room, in the real case, you know, so it's like all these things are already taken care of by the bunch of transformation functions that we already define. So when you, when we actually played them noise that we generated, it does the exact same thing as our simulation. So the Alexa will pick up whatever the Delta we think it will get. And then that Delta is adversarial to its original target of the prediction, which can trick it's all original algorithm. So that's how these things are done 
+
+James Parker (00:31:16) - So much, Billy. Um, can I just ask a, sort of a very basic up question, because you've talked a number of times about risk and security and so on. And so could you just say like in, in a line, what, what it is that you're doing this for is, is this where, when you develop an adversarial audio system, is the purpose, you know, w w what is the w is it about, you know, commercializing selling it to Google, alerting the public as a thought experiment? Just the technical challenge. What's, what's the motivation. Could you just say very briefly what the motivation is behind developing an adversarial system like this? 
+
+Jùnchéng Billy Lì (00:32:00) - Yeah, absolutely. So I was honestly doing at a kind of, you know, because I was familiar, as I said, was this research domain. So I was, and, um, smart audio is, are really picking up in awareness. So I was doing it in more, like, for fun. I was thinking a specific scenario. So like, say like, if NPR plays my, you know, they can do other ways because now we can get to nullify the piece of, uh, Alexa way quartz. So it will not be able to hear the wake word, but in the other case, if I really curate my son, it will really trick the Alexa to wake up with, without you speaking on Alexa, but it's a piece of like, say a highly curated guitar music will be able to wake up Alexa in voluntarily, even if you don't say Alexa, but our adversarial audio will be able to take care of that. 
+
+Jùnchéng Billy Lì (00:32:56) - So the motivation or the original motivation for me was saying, Oh, whatever NPR plays this piece of music. And then geez, like a hundreds of thousands of hundreds and thousands of a household in the us is like, if they are playing their, their MPRs in their living room, there's an Alexa there, all of these will, these devices will wake up. All of a sudden it will kind of be fun to, to think about this type of scenario. And then that is absolutely sort of the things that I kind of, uh, thought about. And then, yeah, it's, it is a huge concern, 
+
+Joel Stern (00:33:37) - Precisely wanted to do the thing, which is the perceived as the biggest threat in the, in, in, in the wrong hands, someone on NPR sort of using the adversarial audio, wouldn't just wake up the Alexis, but then, you know, get, get them to do something, um, and untoward. 
+
+Jùnchéng Billy Lì (00:33:57) - No, no, no, but I mean, that was like part of me, so I didn't really do that. So, I mean, I was thinking that would be like a kind of awareness and for like a non-malicious purpose, as I sat, like, you know, these things will really be a huge risk in a future. If everybody has one of these things that they're home, like really Listening all the Listening all the time, and that will really create an involuntary party of all these devices to do random things, 
+
+James Parker (00:34:28) - You know, the, the field itself of Machine Listening. I mean, just for example, do you even talk about Machine Listening or do you, is that just not a word that you use? Cause I, we we've been using this phrase for various different reasons, but it seems like it crops up in the literature, but there aren't, you know, textbooks or symposia organized around it. So I'm kind of, I'm kind of interested, uh, in that sort of that end of things. Um, but yeah, so, 
+
+Jùnchéng Billy Lì (00:34:53) - So I, I think like, uh, this word is definitely being used by me at least. So it's like, this is a legit word. What else are we supposed to call it? I think like, that's a perfect word for this domain, but, um, as you said, it's, it can be further segmented into automatic speech recognition and then sound event recognition solely because there are different research interest for commercializations.
+
+Jùnchéng Billy Lì (00:35:25) - Some people are more interested in, like some applications are more interested in, in, um, in just purely recognizing the word that you say. And then that is itself a huge task for natural language processing because, um, that itself is its own discipline, but there are more researchers going to the sound events, which are basically environmental sounds of daily events. And these are more of a general like a research studies for how can we better utilize these machine learning models to pick up the daily sound event of not many people as looked into. So these things are more getting more and more usage in say like webcams, like a security applications and Google nest, then you can actually have these sound detection functions activated so that I can pick up dog barks and, um, glass break events and all these security related things. But in my opinion, these, all these things are really getting to becoming like are getting in front of the public eye machines are more and more getting more and more smarter and smarter. 
+
+Jùnchéng Billy Lì (00:36:41) - They're able to hear or understand these sound events. And then of course the speech event that happens around us. So whatever, you know, a lot of people say like, uh, computer vision is privacy intrusive, but in my opinion, the multimodality input or the Sonic input is also privacy intrusive. Because as I said, as if someone gets like 100 hours of your own personal recordings, then you probably get in code, get into deep trouble. If I, or if people like me decide to fake your personal model, then I can fake yourself in a lot of phone calls. And then also detect like, you know, your activities in a sound events. Thanks so much.

content/transcript/dao.md

@@ -0,0 +1,146 @@
+---
+title: "André Dao"
+status: "Auto-transcribed by reduct.video"
+---
+
+André Dao (00:00:00) - So my name is Andre Dao. Um, yeah, I guess I primarily see myself as a writer and as a writer, um, I've been working, um, with them a mix of sort of different forms for, for awhile now, um, chiefly sort of oral history and fiction, um, combined with journalism somewhere in that sort of mix. Um, and so when I was thinking about how that works relates to, um, to your project, um, I suppose there's always been this element of Listening, um, involved in my work as a writer, um, uh, listening and recording. Um, so that begins with sort of my workers in fiction and families, um, is, um, really built around a series of conversations I had with my grandparents over many years using various kinds of recording devices in their, um, in their apartment in Paris. Yes. Um, and, um, I suppose coming out from that, I then I've done our early street work focused around immigration detention on migration, more generally in Australia.  
+
+André Dao (00:01:23) - Um, but yeah, so I guess there's this element of Listening in that work, but there's also this, um, these themes of migration and incarceration. Um, that's both in the migration context, immigration context, but also, um, in my family history, um, there's kind of this, these threads of, um, both being detained, um, in places, um, waiting in places, but also moving across borders. Um, so those are sort of some of the things that I think as float through my work as a writer, um, as a scholar, um, I guess I might've considered myself an extremely junior scholar, um, uh, um, emit about half way through a PhD project at the moment. Um, that is sort of a sensibly on, um, the question of big data and human rights, but fairly early on in that project. Um, I thought I left a lot of work around big data and human rights is, um, looking at kind of what human rights law can do to tame big data or else what big data can do to, um, to make human rights work more efficient. 
+
+André Dao (00:02:42) - And in my PhD project, I've been more interested in, um, what happens to, uh, conception of human rights as well as the work of human rights is, um, as actors within human rights, fear take up the tools and rhetoric of big data, how that all comes together, um, is something that I'm still thinking about. Um, but I think if there's a thread that ties together my work as a writer and artist and that scholarly work, it has something to do with, um, the difference between being a human listener, um, in an oral history context or in a, in a writerly context, um, and, and an Machine scenario and what difference that makes to, um, these various kinds of projects, human rights work, or fiction writing or oral history. 
+
+Sean Dockray (00:03:45) - I know that we're going to spend more time, um, with, on the academic side of things, I think through the interview, but, um, I'm sort of interested just to, to ask a quick question about sort of form and method with, um, your writing and especially where, where you, how you choose to use fiction and when fiction kind of, um, seems to make sense, particularly when you're sort of balancing or moving between, uh, forms of oral history and fiction, is it to fill gaps? Is it to, you know what I mean? Is it, um, like what, what, what is it that fiction offers you in the writing process? Hmm. 
+
+André Dao (00:04:25) - Um, there's definitely partly a, um, a slightly practical drive to fiction or at least in terms of, so the context of, um, my fiction writing is it is family kind of history project. Um, and yeah, there's definitely a filling gaps, um, part to, to the turn to fiction, um, and also, um, um, some sense of, um, wanting to, um,
+
+André Dao (00:04:58) - I have a kind of almost plausible deniability when it comes to my relationship with my family. And when this book gets published, I'm publishing it as fiction, I think makes a difference. I think more in as a matter of like artistic form though. Um, I also, I think that, um, for my writing, working in fiction, uh, allows me to focus more squarely on, um, I suppose the, the work of memory as kind of a fictional process, um, in itself. Um, but when I say, yeah, so when I say that I'm working in fiction, I think there are lots of non-fiction writers that I admire greatly who use very similar techniques. They happen to batch it or their publishers batch it as non-fiction experimental nonfictional, so on, but it's in that kind of, um, realm of so that some of them thinking of, um, for example, this Maria to mock and, um, as a non-fiction writer, but, um, she uses a great deal of techniques drawn from, um, not strictly from non-fiction. 
+
+James Parker (00:06:12) - Should we maybe move back onto global pulse, uh, and the work on big data? I mean, we, we, we could spend a lot long time talking about, um, writing more generally, but you've been doing some work on this thing called the UN global polls. I was utterly amazed to find out, you know, that this existed and I'm sure that many other people would be too. Um, can you explain a little bit about what it is, where it comes from? Um, what, what kinds of projects are undertaken in its name?  
+
+André Dao (00:06:43) - Sure. So, um, the global pulse is, um, badged as the UN secretary General's innovation initiative and essentially, um, what that means is, um, it's the UN's attempt to, um, get to grips with, um, what they understand is big data technologies. Um, so largely machine learning based on machine learning and to try and harness that technology to, um, to complete the, um, military chief, their human rights and development mandates, um, its headquarters, um, is in the, um, is in New York, um, just a few blocks down from the UN headquarters. Um, but they have set up two satellite labs. So one in Uganda and one in Indonesia, and they basically produce, um, prototypes of big data tools for development, um, data gathering tools. So they might partner with, um, other UN agencies like a UNH UNHCRs to say, monitor Twitter feeds, um, uh, of, um, to pick up data about, um, migration routes, um, through the med. 
+
+André Dao (00:08:01) - Then they might also work with, say the Ugandan government to monitor, um, Ugandan talk back radio to try and, um, pick up early warnings of hate speech and potential for, um, for violence against refugees in Uganda. Um, or they might use big data techniques to monitor and predict, um, smoke, um, and pollution air pollution in Jakarta, I guess the impetus behind, um, the global pulse seems to be the UN anxiety about its own competency to complete its work. So I think one of the kind of hidden hidden stories to the global pulse is the UN's ongoing financial crisis. It's just run out of money. And it, I think it looks across at big tech and, and sees an opportunity to continue to do its work in a way that, um, I think they, they, they both see it as a way of making their work more efficient, but also by taking up the rhetoric around big data, that it can attract, um, a certain kind of investment, whether it's from the technology companies themselves or from governments, um, particularly in say Scandinavian countries that are very excited about supporting data projects in places like Uganda and Indonesia. 
+
+André Dao (00:09:29) - And I think that side of the story becomes pretty clear when you look at the launch of the global pulse, um, back in, I think 2012, um, where, um, the, then secretary general ban Ki-moon launched the pulse alongside, um, you know, former executives from Amazon and Apple. Um, and it, the global pulse is headed up by, um, a former head of Microsoft's humanitarian systems. So it's.
+
+André Dao (00:09:58) - It seems pretty clear. I think even just from the personnel involved that, uh, this is the UN's attempt to, to ride, um, uh, kind of the, that big tick way. 
+
+James Parker (00:10:11) - Could you maybe give us an example of how one of these experiments or tools works? I mean, I was particularly interested in talking through, um, I think what they call the radio content analysis tool, just in terms of thinking about Machine Listening or how Machine Listening might be deployed by or produced by something like the UN what is, what is this radio content analysis tool? 
+
+André Dao (00:10:41) - So that's one of the prototypes before I explain the radio content analysis to it. Um, I did want to mention that one odd thing about the global pulse is in my research so far. Um, all of these, whether it's the radio content analysis tool, any of the other, um, um, initial, um, kind of programs that they set up, it's, it's never clear to what extent any of it gets taken up by other parts of the UN. Um, so they've been operating for kind of, um, eight years now, and it seems that they kind of, um, continually produce prototypes and push for more use of machine learning techniques throughout the UN. But the actual tools themselves don't necessarily seem to have a lot of take ups or the importance or the significance of the club post seems to lie elsewhere in the actual use of this technology, at least at the moment, um, which is maybe something we can come back to the radio content analysis tool is, um, one of the prototypes developed by the Kampala lab in Uganda. 
+
+André Dao (00:11:46) - Um, and it's, um, function is to essentially analyze, um, Ugandan talk back radio. And so in the documentation about the tool, um, the pulse really emphasize this kind of overwhelming amount of, um, audio, um, that's produced by Ugandan radio every day. Um, and the point there is to really emphasize the, the incapacity of the human, um, in the human listener in that situation. There's no, um, there's no possibility of having enough people listening to enough radio stations to capture everything that's going on. Uh, and so that's where machine learning comes in. And so the tool, um, works by first filtering out music and ads from all of the radio, played it in Uganda, um, and then producing, um, using speech recognition software to produce transcripts of, um, of the talkback radio portions. Um, and then it uses keyword filtering. That's been, um, kind of trained by by human analysts and human tag is to, um, to pick out supposedly relevant snippets of conversation. 
+
+André Dao (00:13:02) - And so that the diagram that the posts uses to illustrate this processes, you kind of start with this really large box of, um, structured, um, data. And then that box gets smaller and smaller as I, um, as I intervened with more machine learning processes, combinating in this kind of useful tagged and categorized information at the end, uh, and that useful information at the end, um, various between it could be their examples that they give is it could be, um, picking up instances of hate speech on, on Ugandan radio again, um, you know, aimed at vulnerable groups like refugees. Um, but it could also be more in the vein of kind of user feedback on, on service services. So refugees calling in to complain about, um, say the health and sanitation services in a refugee camp. And so that's the basic idea of it. So 
+
+James Parker (00:14:00) - I should, you know, play my card is in a way as, as somebody who works in a law faculty and I just, what you were describing then I think it sounds intuitively plausible on some level, but as somebody who's read hundreds and hundreds of pages of judgements at the end of extremely long trials, debating whether or not something constitutes hate speech, I'm just immediately thinking, how on earth? What, what, who, what kind of things are being picked up on, uh, is it, you know, just references to ethnicities? Is it like, what, what, what is constituting hate speech for the purposes that analytic or whether or not it works? I just want, I'd just be fascinated to know what are they even thinking. It would mean for a content analysis to, to be able to detect something like hate speech.
+
+André Dao (00:14:58) - Yeah. So from what I can say from their documentation, the, um, I mean the process, uh, isn't necessarily reductive, um, and the ability to, um, to tag, um, sort of 30 seconds snippets is potentially hate speeches. Um, it's pretty simplistic. Um, and the sort of K-12 filtering that they use it. Yeah, it is. I think it does say that it's as simple as in the end, the team of human designers, um, coming up with lists of, um, of words that seem to be relevant, um, to say hate speech. Yes. 
+
+James Parker (00:15:48) - Sorry. Sorry. Um, and those teams of designers are in New York, right? So that's not a team of designers that for example knows the euphemisms in the local language. I mean, I'm just thinking of the Rwanda genocide and the specific ways in which you from Islam were deployed in the promotion of hate speech there. Right. So I don't know how an algorithm would be able to detect that, so that it's New York designers who are inputting that, those analytics, 
+
+André Dao (00:16:17) - Um, at least at the first instance, I guess, um, the promise of the, um, of the tool is that, um, as it gets rolled out, those things get refined, but in some ways that's, I mean, your question, I think, and, and particularly a reference to say that the, kind of the legal procedures involved, um, posts Rwandan genocide, I mean, it's kind of precisely that sort of lengthy convoluted, legal proceeding that, um, this entire, um, like the, the, the pulses sort of raising for bang, I mean is to sort of short circuit that right. Is to say, and this is in, uh, in, in all the ways that they describe the aggression of the tool is that you're dealing in situation of crisis. Um, that there's a vast amount of data out there that, um, is described as data exhausts. So you get this picture of it, a being produced for free, or it's sort of produced as detritus following people's, um, kind of everyday lives. But particularly I think that image of exhaust is it's escaping, it needs to be captured quickly. And that combined with this always the emphasis of crisis, um, and always the emphasis on real time data. Um, the point is you don't have the time for those sorts of legal niceties, um, or even social cultural understandings of euphemisms and so on. We just don't have the time for it. And the UN needs to kind of act almost instantly. So that's where this technology comes in. 
+
+James Parker (00:18:00) - So it's trading out preemption, well, pre trading out judgment and justice mechanisms for a logic of preemption and prevention. 
+
+André Dao (00:18:10) - Absolutely. Yeah. Prevention and yeah. So it's not, it doesn't really, um, operate under a rubric of justice in that sense. Um, it's much more something like executive role or, or management. Yeah. 
+
+Sean Dockray (00:18:27) - Can't help, but also think of the financial sort of problems that you referenced at the beginning of, um, of, of all of this and that it's just imagining that tools like this would also sort of be imagined to be cost, maybe revenue generating on the one hand, but also cost saving on the other hand of, um, you know, because then in theory you need fewer people on the ground, like who have this kind of, whatever deep knowledge of, um, kind of meanings of language and all this kind of thing. And instead you have a system that can just sort of alert the right people to potential problems. So, um, I think that's interesting tying, tying the development of this tool to the kind of financial problems at the UN more generally is, is experiencing. Yeah, 
+
+André Dao (00:19:20) - I guess what I mean, it's interesting that one of the things I'm interested there is that, um, so while kind of expert the rule of experts that, um, uh, an international organization, like the UN has been, it has been criticized, I think for, um, bringing in, um, to both the human rights and development.
+
+André Dao (00:19:43) - So to see is right, is like you bring in these experts and, um, they can, um, produce this kind of technocratic rule. It seems like, um, what the policy is doing goes beyond, or even sits outside of that logic, because as you say, it's, um, even bypassing that expert with the kind of data, you know, local knowledge or history or, and moving towards a model where yeah. Someone, um, sitting in Manhattan, um, is not even necessarily being the best of the monitors, but they just simply the person that designed the algorithm that the algorithm then does the monitoring. 
+
+Sean Dockray (00:20:28) - James are a little earlier, I sort of said slipped and, or on the way to saying something else said who who's writing this code, you know, and, and, uh, a few in a few of our conversations, there's been talk about getting who is at the table at the conception and design of these systems. I'm just sort of interested. Do you have a sense of, of, um, who, well, on the one hand who's sort of conceiving of the systems, but also who's, who's actually implementing them, who's working on what types of people are being employed to actually develop these systems? 
+
+André Dao (00:21:02) - Yeah. So it's probably important that to acknowledge that. So while the is headquartered in, um, New York, um, from what I can, can tell the, like the Kampala lab and the Jakarta lab, um, to employ a lot of local stuff, um, and you know, they are, um, employing local data scientists. Um, I guess the question there is for me is, um, what, and what I haven't, um, yet, um, sort of followed up on is, you know, so where were these data scientists try and cause kind of, I guess on a very surface level, you look at, um, you know, the fo the staff photos for those two labs. And it seems like, you know, it is, um, say a lot of Ugandans working on the radio content analysis tool, uh, on some level. But, um, yeah, there's still, uh, a question in my mind of, well, you know, what kind of training, where were they trained? 
+
+André Dao (00:22:08) - What kind of assumptions about, um, about data? Do they bring to the project, um, on the question of who, uh, another way that I've been thinking about about that is not so much at the design level, but, um, I guess one of the things that the post says about the radio content analysis tool and about their other machine learning tools, um, is that if it's within this rubric of, um, uh, participation. So the UN for, for a long time now has been, um, talking about, uh, people centered development, um, there's participation and informed consent. Um, that sort of language is incorporated into a lot of their human rights work into their development work. And what that means seems to be shifting at least when it comes to the involvement of the global pulse. So participation in a process in a kind of human rights process, um, it seems to be participation could just mean that your data is being scraped from a Twitter fade or that you're, um, or that some people's, uh, calls to talk back radio in Uganda have been, uh, tagged and categorized and turned into a chat and read in, um, at UN headquarters in, in York that that's a form of participation or that that's what makes development people centered. 
+
+André Dao (00:23:51) - And so that, um, yeah, I find that a very interesting potential shift. 
+
+James Parker (00:23:58) - I mean, it's also a form of doublespeak, isn't it? You mean, you wouldn't say that, you know, counter-terrorism, you know, is people centered, for example, you know, w w w I mean, I'm really struck by how similar, what you're describing is to many of the, um, NSA's programs that were, you know, revealed by Snowden. And in fact, you know, around precisely around the time when global policy is being developed, there's a quotation in the, the essay that you shared with us, but blew my mind. Um, so this is the independent expert advisory group on the data revolution for sustainable development, which I kind of want to know more about, but you quote them as saying.
+
+James Parker (00:24:43) - Never again, should it be possible to say, we didn't know, no one should be invisible. This is the world. We want a world that counts. So a world in which we count everyone all of the time. And that is exactly the same rhetoric as the NSA is program of total capture in the name of counter-terrorism. And one of the things that interests me is how it's possible for an organization like the UN to speak that way without, without recognizing the in congruency, like, you know, um, it seems like there must be a lot of in there must be a lot of, is it a lot of faith in the idea of data for good, or just a fundamental commitment to them not being, you know, the NSA so that, so that it doesn't, it simply doesn't present a problem, but I was just, I was just amazed by the, kind of the hubris of that as a goal for the UN. And I guess, I guess that site sort of slides into a question about what the UN itself understands to be the risks of the programs that the big data turn, or what have you that is developing and how you think alternatively, uh, we might think about those risks or, um, or what have you. 
+
+André Dao (00:26:06) - Yeah, it's quite odd, isn't it? That, um, surveillance as a word and as a concept, just doesn't seem to get, um, picked up in this conversation, like as soon as you're in that kind of data or AI for good world, um, this, yeah, quite similar seeming practices just don't get that label surveillance, um, and seems to be that surveillance has some tie to security, um, and that we have this kind of conceptual link between those two, um, and that somehow in the kind of AI for good realm, because it's what, where it's not explicitly based around security or law enforcement, it's no longer surveillance in that quote that you read out about, um, the world counts. Um, so I think what, what I found interesting in my research has been the move that, um, the UN makes the move to that independent advisory group on the data revolution, um, makes, uh, to authorize this vision of a world in which everyone, um, is surveilled essentially, but they, they can't is they, um, they talk about the right to be counted. 
+
+André Dao (00:27:29) - And so actually this is where, um, it's, I think a good example of the use of human rights, at least conceptually to authorize a move that otherwise we'd think of in of surveillance and, and, and, and, and certain kinds of people would then, um, you know, have a repertoire of resistance to, um, but by using this move of saying, actually there's a right to be counted in. So that, that right to be counted, that is to say that it's already a human rights violation to not be counted by these processes, um, to not be saying or heard by Machine Listening or Machine vision. Um, I think that's just such a interesting move because it's very, I think in, in sort of a lot of the languages available to us for, um, political activism, it's quite hard to, um, Dave and say to that moves happened first of all, but then to, um, kind of intelligibly, um, articulate a resistance to this idea of, um, so they, they, they placed the right to be counted as prior to every other human rights. So you can't have, you know, the right to ha you can't have your right to housing fulfilled unless you've first been counted as a person. Um, and there's this kind of conflation, I think, between, I guess, a deliberate confession really of, between this idea of being counted or being measured and being quantified through algorithmic processes and counting as in, you know, mattering, but it's in that. Yeah. So I think that's in that sort of slippage. Um, it then authorizes a whole lot of stuff that we might otherwise have concerns about. 
+
+Joel Stern (00:29:30) - I would say it's almost, um, sort of analogous to like, um, the right to vote or something in a, in a kind of democracy like to be counted is sort of to participate in the functioning.Oh, of a sort of political system in the same way that, you know, perhaps, um, voting in, it might be, it's almost like a, uh, sort of more passive way of having your say let's, let's say. And, um, I kind of want to, um, you know, ask a couple of like bigger sort of DAMA questions too, to sort of provoke let's, let's say, um, for, from you, uh, sort of let's say even a provisional sort of position on, um, this policy. So, I mean, I'm tend to, attempted to ask, you know, do you think that this program pals or, or certain sort of tools just, um, simply should not exist, you know, should, you know, be, be sort of put to a stop, uh, likely to do more harm than good et cetera. And, you know, so I, I, you sort of fundamentally skeptical about the use claims they make and, you know, if so, w w what are the, um, you know, negative horizons, like what, what are the sort of most, um, serious problems and concerns, um, that, that we should be mindful of? 
+
+André Dao (00:31:08) - Hmm. Um, so am I fundamentally opposed to, um, to the use of these technologies in a human rights context or, or by the UN? 
+
+Joel Stern (00:31:20) - Sure. I mean, H however you want to interpret that, but I mean, I think Pat perhaps narrowing the question to, you know, the, the UN and this sort of specific program when, as you're sort of researching it and looking at it and going into it sort of more deeply a thinking this simply sort of shouldn't exist as a program, or it should be fundamentally different to what it is, I guess I'm just trying to get to the sense of whether what we're saying is these programs should be better sort of safer, more regulated, more sort of stringently sort of applied, or are we saying, no, this is a fundamental wrong turn at the end, we should sort of what, what's the kind of, um, from, from an activist or a sort of political intervention, um, into this, what what's the position that you think is, is going to be the productive one? 
+
+André Dao (00:32:19) - So maybe my way of grappling with that question, um, is through, uh, I, I know that in the, um, kind of the, the prep doc that, um, that you guys sent around, those, the question around how I, why I use, um, jurisdictional thinking and in my PhD project, and maybe a very quick, um, answer to that release of hash Lance to what, what I useful in jurisdictional thinking is, um, is emphasis or, um, yeah, that part of jurisdictional thinking, I think, um, emphasizes a question, uh, who question of like, kind of who decides in relation to your question, Joel, then around these technologies, um, I've been wondering about whether or not that who question does really matter. Um, and the reason I'm thinking of that is sometimes I say artists and activists taking out some of these technologies, right. Um, to do quite interesting work, quite useful work, it seems. Um, and so I'm wary of maybe accepting fully the, um, kind of the input of your question and saying, because while I am quite skeptical of the use of these technologies by the UN, I'm not sure yet whether it's a question of the technologies or what yeah. Does that make sense of, I've been wondering what, what difference it makes who's wielding the technologies and for what purpose, uh, 
+
+James Parker (00:34:06) - Could I follow up there? Because, because it seems to me that the question is not about in your writing, the question is not about the technology. You talk about big data as a concept that has been taken up by the UN as a program that's being taken up by the UN, not a technology that's being taken up by the UN. And so when you start to think that way about big data as an orientation to the world, which has a certain kind of visual rhetoric and buys into a language of innovation and can start to imagine something like the right to be counted, then.
+
+James Parker (00:34:48) - You've got a much more sort of complex, you know, socio-technical ideological beast than, than simply a technology which can be deployed or not deployed by this organization or person that one, you mean? So what if we were to reframe this question and say, are you opposed to the way in which big data has been taken up by the UN, but as opposed to asking, are you, are you opposed to gods or data analytics or machine learning per se? 
+
+André Dao (00:35:26) - Yeah. Um, yeah, I mean, you're right. As a program, um, what has happened with UN takes up big data as a program is this vision, um, of where, um, I think what this one, the global pulses, chief data scientist, um, talks about by, by 2030 we'll know everything from everyone. So no one will be left behind. Um, so it's in so far as that kind of logic is carried with this program, with these technologies of the knowing everything from everyone that's, um, that's sort of, yeah. Um, certainly opposed to, to the UN moving in that direction in terms of the, that the negative horizons of things we should be worried about. Um, and this ties back to something that James, you asked, um, a little earlier around the risks, the rest of the UN SES is, is very squarely around the privacy question. And I think that's, um, probably unsurprising, um, to you guys that they say the only potential problems around say the radio content analysis tool or any of their other pulses of, um, projects is kind of individual privacy. 
+
+André Dao (00:36:58) - Um, whereas I think there are much bigger, um, and very different types of risks involved here. I mean, one of them we've alluded to before is, um, the global pulse, um, and the taking up of big data by the UN is, um, proving to be a way into international institutions for big tech. Um, so, um, that has, it seems to have an interesting effect on the kind of authority of the big tech companies themselves as they work with the UN and do human rights work. Um, I I've seen the, um, the CEO of Microsoft kind of explicitly, um, and quiet. I think Kressley make these connection way. He talks up Microsoft's human rights work and its partnerships with the UN and then, um, sort of immediately segues into their facial recognition software and how they're doing all the right things around privacy of facial recognition software. But you know, what the audience is left with is this initial impression of, but the Microsoft is working with the UN it's doing this good human rights work. I think that already does some authorizing work for the facial recognition software. For example, 
+
+James Parker (00:38:35) - Particularly if, um, the programs aren't even being an enacted so that they're doing good human rights work remains entirely at the level of speculation. It's enough for it never to be enacted so long as the collaboration is there. I mean, it seems to me that the, to invent such a thing as the right to be counted, even if it's very early days, is doing a lot of work for big tech, you potentially down the road as well, a complete reconfiguration of what it means to be in the world to be a subject of development. I mean, that's sort of a little bit far away, cause it's not exactly like a massive rights document that's been fully elaborated, but even to see that idea is, Hmm, 
+
+André Dao (00:39:24) - Well, they, um, yeah, the, one of the other interesting, um, kind of images from the same report that introduces this rat to be counted is, and that also talks about that. You know, the, the world that we want is the world where everyone counts, um, is this image of two worlds. They, they talk about, you know, there's, there's a world where, um, we have great access to the internet and.
+
+André Dao (00:39:52) - Great access to, to these new technologies. Um, and there's another world where, you know, people don't have that. And I think that framing immediately puts us into that kind of familiar developmental timeline of that, of, you know, from the pre-vet, from the primitive to the advanced and what it does. And he just, it creates a single image of the future that we're all sort of striving towards, which is where these technologies are fully rolled out. And this program of knowing everything from everyone is, is fully realized. It kind of presents the question only as how do we help Africa and Asia and the developing world catch up to that world where we know everything about everyone, and it's not, you know, it obviously obscures the question of whether or not that well is when we want to get to in the first place. Um, and it presents that gap as a, as a question of justice inequity. Um, and so if you're concerned about questions of social justice, you focus on the gap and you miss the fact that, you know, what, what purportedly aiming towards has, um, potentially huge negative ramifications. 
+
+Joel Stern (00:41:11) - I mean, it begs the question of, um, like I'm just sort of thinking about a statement like everybody counts from the UN in relation to a statement like, um, black lives matter, for instance, like if they're sort of somehow not notionally doing similar kind of work in trying to say, you know, um, create a sort of politics of, of sort of who is counted. And, and I'm just wondering about like the UN trying to roll out a program like the radio analysis tool, for instance, in the United States or, or in Australia or in the UK, um, and how different programmatically and politically it would be as it sort of intersected with all of the sort of infrastructures that exist in those countries. Um, and so, so how much the kind of relative, you know, political power of, for instance, the United States and the UN in relation to Uganda structures, a pro a program like this, if we sort of imagine this tool being rolled out in Australia across sort of hundreds of radio stations, it's, I mean, it's very speculative, but I'm imagining there would be a number of, sort of, it would just be called easier. 
+
+Joel Stern (00:42:30) - Well, that, I mean, that's also goes back to the point you're making about the NSA, um, and the fact that, um, these programs essentially already exist in other sort of departmental and political spaces. And so they are kind of then taken up in, uh, by the UN and, 
+
+Joel Stern (00:42:51) - And perhaps in a slightly more transparent, more transparent than the NSA imaginally 
+
+André Dao (00:42:58) - Is this kind of service delivery aspect to it. I think which, I mean, it's interesting for one thing is the question of whether or not that's how the, this positioning of the UN as you're delivering services to customers. And actually there's a, um, a slip of the, of the tongue at the launch of the global pulse, perhaps because, um, it's not the secretary general, but I think an under an under secretary, but he refers to essentially, you know, the global poor as the UN's customers and that this, this, that the post is going to help us know more about our customers. Um, and perhaps that's because he's following up from, or introducing this, you know, execs from Amazon and Apple, but that's already a shift in how we've, uh, of how the UN would view it's. Um, so I that's a shift in how the UN presents its work, but, um, Bay, I think, yeah, it just, it fits within a much more benign, uh, kind of understanding of what surveillance is if it's kind of badged is, um, yeah. And, and consumer feedback really, you know, um, even though ultimately what some of this data could and probably will authorize will be international intervention, military intervention, that's that's, I mean, that's certainly where this idea of early warning systems around hate speech. I mean, what, what can that be pointing towards except a situation in which, you know, Machine Listening picks up enough data on, um, future high crimes that authorize his military intervention. 
+
+James Parker (00:44:44) - I mean, that, it can't be a coincidence that they've gone for radio first. I mean, you know, after the Rwanda genocide, where the story that gets told is that radio is one of the prime drivers and so much of the discourse after it was either we should have intervened early. We being the U S and Europe primarily.Oh, we should have jammed the radio stations at the very least. It just, it just must be that it was conceived with that in mind. I mean, is, is there other explicit about that? Yeah. 
+
+André Dao (00:45:13) - Make that, um, that connection explicitly, but yeah, I mean, it, it's definitely floating around in there, I think. 
+
+Joel Stern (00:45:22) - Mm. I mean, they give lots of, they give lots of examples to do with sort of, um, you know, environmental disasters, like, like, you know, floods or, uh, and sort of the way in which responses to that, to those might be sort of ex accelerated. But I suppose again there where we're in territory, uh, where, you know, what happens in the wake of an environmental disaster is not always, um, you know, positive development, but also kind of exploitation of that disaster, especially in relation to sort of, you know, multinational companies. I just had a, um, uh, I just wanted to return to this, um, concept of a world that counts and everyone counting and just to make a really dumb connection to the, um, uh, to the census or to a kind of like census within the country. And that's obviously like often rolled out with, with exactly the same justification, which is, and sometimes with the same branding as well. 
+
+Sean Dockray (00:46:30) - We, you know, we want everyone to count and, and certainly in the U S there's like, uh, there's a lot of political struggle around the census and, you know, we'd, we do want to count everyone, um, particularly undocumented workers, because that's how political power is apportioned. That's how, you know, tax dollars and funding, you know, there are a lot of, so I can see the, the UN sort of deploying that, um, kind of rhetoric and justification, but I guess the big difference. And so what I'm asking you to do is a little bit too, cause I'm naive as to explain the UN to me, the part of the difference is that the, like what gets apportioned in the UN's vision of a world in which everyone counts, because like through big data, like what, um, in a way it seems like part of the problem is all of this is collected, but there's what mechanisms are there to take it up and use it in a, in a, in a, um, positive way. And maybe that's kind of related to the problem of a lot of these, uh, observation that a lot of these pilot programs aren't actually taken up, because what, um, maybe the question is like who or what areas within the UN have the capacity to, um, make use of these, um, these technologies, you know, in, in a, in a, in a useful way. 
+
+André Dao (00:47:53) - Hmm. I mean, that's, I think that's why the, the slip of the tongue where the, um, with a UN official refers to, um, the global portrait is the UN's customers. I mean, it's, it's a revealing one, not just because of the, of that, leading to that, um, question around corporate power, um, and so on. But, um, I think it's also interesting simply because it's not correct, it's the UN is made up of it's kind of constituent state members, and this what's really one of the really fascinating aspects of everything that the post does that tries, I think, to reframe or it's part of it's part of a larger, ongoing process of within the UN of reframing it's its constituency as the peoples of the world, and that it derives some kind of direct authority to, um, to make better the lives of, of all the peoples of the world. 
+
+André Dao (00:48:57) - And I think that is a really important part of the function that these technologies or this program plays for the UN to be able to position itself as being both capable of and authorized to, um, intervene positively for people, um, you know, as a, as opposed to having to work through States, nation States. And so that, yeah, I think there's a really fascinating, um, thing going on there around competency and perception of competency. Um, so, and that's why the, the lack of uptake of these technologies and projects is so interesting because it seems to be that you can presenting data is as good as.
+
+André Dao (00:49:54) - Having the actual capacity to change the vein. Um, and maybe an interesting image that captures some of this, um, comes from the document around the radio content analysis tool. Um, they illustrated by these really strange images, um, of Ugandan women standing in kind of, um, empty fields, holding radios up against their faces. 
+
+James Parker (00:50:25) - I'm on the, um, the UN global pulse website right now. And it's, it's the landing image, right? You sort of, you, you land there, it says big data and artificial intelligence, sustainable, uh, humanitarian, yada, yada. And then on the left-hand side, there's a woman with a portable radio held up to her face in a field. 
+
+André Dao (00:50:45) - And so it's a strange image, right? Because, um, I mean, I think your first impression is it, it, it fits perfectly well with this, um, idea of participation of, of the UN creating some kind of a platform for, um, the peoples of the world to speak directly to the UN. And that's what makes the UN competent to intervene on their behalf. Um, but of course she's holding a radio, not about fun and so already in the CMH, there's this? Um, yeah. So that's the image that I'm thinking of there's already, um, something we had going on in their own kind of marketing materials about, you know, supposedly showing this participation. Yeah. So it's a, it's an image that I keep returning to in my, in my writing on the post. So 
+
+Joel Stern (00:51:48) - Is it, does it, it sort of speaks to Machine Listening in such a weird way because it's sort of, it's so far from a kind of computational algorithmic sort of data gathering kind of image. It's so much about community and a very old fashioned media device, and there's also some really weird displacements going on, right? So who's the person doing it, Listening in the image. It's a Ugandan woman who are the people doing the Listening in the UN global posts? Well, no one, but if anyone in Machine that's been programmed by, you know, ultimately all roads lead back to New York. So there's a lot of, a lot of political work in that displacement, the kind of attempt to convince people that the Listening is being done. Presumably the idea is that it's being done on behalf of the Uganda woman. So we, we, we might as well be Listening for you in Oxford, a role as, you know, benevolent surveillance that's right. 
+
+Joel Stern (00:52:53) - Because she's listening to talk back, um, and radio and therefore, so, so might way, uh, in order to sort of, um, understand on a macro sort of level what Shea is understanding, just sort of with the one radio, I don't know, but the ability, she can't understand it from the one radio because she's not listening to all of it all the time. So it's an enhanced, it's better. The UN can do much better like that than she ever could. I mean, it's so it's so strange. I mean, it seems patronizing to me as well, quite frankly. 
+
+André Dao (00:53:29) - Um, there's I think, uh, a blog post, um, in which the, the idea of the right to be counted kind of gets introduced by two of the experts on that advisory groups that I mentioned before. Um, but in that they asked a question, um, can they take your voice, which I think, um, kind of nately, uh, intersects with the CMH. Um, but also is it the entire program, um, which they, uh, you know, I think clearly the posts, um, and so that in the affirmative that, that here we have data giving voice even where voice is impossible because she's holding the wrong device. Um, and yeah, uh, recently I was rereading, um, uh, Gayatri, SPI VAX article, um, can, can this whole Bolton speak and, um, it kind of, yeah, the CMH and, um, that question about data giving voice reminds me of the, the conflation between two types of representation that Spivak talks about in that, um, in that essay. So, you know, between representation is something bang, symbolize something, paying big shit over, represented in that sense, and then represents it as you know, that idea of speaking for in the way that we're represented by parliamentarians.
+
+André Dao (00:54:58) - Um, and that those two things go run together. Um, particularly in the English language. I think she says in German, there's actually two different terms for those two meanings of representation, but yeah, there is something happening here I think, was the posts where the ability to represent someone through data, um, older ability ability to represent many someones for data, um, same to then also immediately carry, carry with it. Um, the authorization to represent someone as in speak for them act on their behalf. 
+
+Joel Stern (00:55:39) - I'm wondering if, um, the, cause I think, um, we were gonna need to finish up in a, in a, in a few minutes, but just I'm thinking, you know, where we've arrived at is reef is really amazing and it could be, could be interesting, um, to have you sort of depart from this image and kind of pre present sort of an, an analysis of what is the work that this image is sort of doing, um, in its representation of Listening against the form of Listening that is sort of happening with the radio analysis tool and maybe those two questions, um, can data give voice against sort of spear of X quick question, can the subaltern spake, you know, it just could be a really nice way of, um, sort of framing this image. I mean, uh, Andre has, um, given many presentations enrich that's more or less exactly what it does. 
+
+James Parker (00:56:39) - I'm sorry. Yeah. Okay. Um, but I think it's a great idea to, I mean, I think, you know, I don't know if we're rounding out the conversation, um, but there is, there remains the question of the live context to unsound and whether, I mean, I think with that, because just with the unsound context, the connection has to be strongly to forms of Listening, um, rather than to sort of let's say the politics of big data more generally. So I think in that context it would, it would sort of be to depart from an image of Listening, which, um, yeah, anyway, I mean, should we, should we sort of end the sort of formal bit of the interview? Um, I mean the only other things I was going to ask were about COVID and, uh, the interception of migration, uh, distress signals from boats, carrying migrants and refugees, it seems like that, you know, are kind of part of this expansion, you know, like everybody's responding to COVID. So the UN got its big data COVID thing, techno solutionism as, uh, as a way of responding to COVID. Um, and that, you know, the mounting refugee crisis is just another one. Um, so I just wondered if you, any, any particular thoughts on, on those either of those projects? 
+
+André Dao (00:58:01) - Hmm. Yes. I haven't, um, been following the COVID, um, part of, um, out of the pulse particularly closely, um, except yeah, I guess it just shows the, I guess the, the logic of the pulse as these kind of innovation lab is that they're in the same way that they, um, there's this language of, um, matching up innovators with problem, uh, problem owners. And so they're, they're the innovators and they're looking for the UN's problem owners. And so when it was a refugee crisis, it's the UNX here, who are the problem owners they own, they own the problem with the refugee crisis, I guess, which sets us that. Yeah, I think that they kind of by the very like kind of constituents that they have to move from crisis to crisis. And so when they say the COVID like it it's inevitable that, um, this innovation initiative has to show the worth of its work and its program in relation to whatever the, the latest processes, 
+
+Joel Stern (00:59:21) - But it's so striking the way that they lack the language and that, that move exactly mirrors, everything that every tech startup and major, uh, you know, Google, Amazon, whatever did, I mean, it's just suddenly tech presents itself as the solution to the problem. And, and, and it's an, a further opportunity for expansion 
+
+André Dao (00:59:44) - Actually. Um, the one thing that I didn't mention that, um, is probably if there's sort of an end point.
+
+André Dao (00:59:52) - Uh, or a hunch that I'm going on with my PhD project it's in relation to, um, the pulses, the, the one project that sort of seems to underpin all of these other more prototype, um, style projects, is there push for a kind of global data comments and, and push for what they call data philanthropy. Um, and, and that is essentially there. Um, it seems that that they present these tools like the radio content analysis tool, um, to, um, particularly to large corporations that hold a lot of data on people. Um, and so this is the work that the UN can do with data. If you see the value and the worthwhileness in that work, you should donate the donate that donate the data that you hold on your customers or whatever data you hold, um, to the UN, um, so that we can start to turn these tools, whatever tools we have for analyzing data, um, onto your data sets. Um, and, and that the vision there is of no longer having data in any silos, any corporate or national silos, but a big, um, global data commons. Um, and so that seems to be one of the primary purposes of, uh, of these individual projects is to, um, to convince the holders of data to, to, to share it in this big happy kind of weld, if 
+
+James Parker (01:01:32) - True data comments, or a data commons that is possessed by the UN. Right. 
+
+André Dao (01:01:36) - My understanding is that it would be open for old, that wished to do good with AI. 
+
+James Parker (01:01:43) - Well, that, that that's not going to happen. Is it, I mean, why, why would, why would Google give up? Yeah, 
+
+Joel Stern (01:01:49) - I mean, that's it because if, um, Google and Facebook handed over their data to an open comments, then they would all just grab each other's data. But I mean, it's from one another, but I mean, it would be good. It'd be good. And then we, and then we could just sort of rebuild a not-for-profit Facebook out of, I mean, good might be a strong, a strong word to use. Okay. Be better than what we currently have possibly 
+
+André Dao (01:02:17) - And thought about it from, from that, from the perspective of, of your Googles and, and why they exist. I, um, had mostly been thinking about it from, uh, from the post his perspective, but that's yeah, that's interesting point.

content/transcript/joler.md

@@ -0,0 +1,132 @@
+---
+title: "Vladan Joler"
+status: "Auto-transcribed by reduct.video"
+---
+
+James Parker (00:00:00) - I th I think it might be, it might be worth just beginning at the beginning. Vladan uh, and, uh, asking you to introduce yourself again and say a little bit about your, you know, your background and your work. 
+
+Vladan Joler (00:00:13) - Okay. My name is Vladan. I'm based in Serbia. So, uh, yeah, I think this is kind of explanation of, uh, of like attain visibility infrastructure started bitten me, like maybe deeply like five, six years ago. So we were running one organization called shared foundation and organizing some events for these kind of stuff. But the moment we realized that you cannot solve a lot of problems just with like big meeting, interesting people, because we understood that in moments when something was like captain in Q and the region, we didn't have the capacity to react in a way, like more like expert capacity in sense of like, you know, legal analyses, technical analysis, those kinds of stuff. So we started to do, to organize one, like expert group of people and, uh, uh, basically bunch of lawyers, experts, cyber forensics, and song. And this is one moment, uh, I started to do some kind of like, really like personal investigation of like, because I always wanted to, it was always impressed with this kind of internet maps back in like early two thousands and stuff like this.  
+
+Vladan Joler (00:01:54) - And, uh, so I realized that I'm now able to finally to draw those kinds of maps, uh, network maps, internet maps, and stuff like this. And then, you know, with the first map, you know, when you are able to see something, how it looks like, then you are more questions are coming. What are those points of centralization things that I'm seeing in the map? And this is when I basically started to shape my research in some kind of call off, but on the way with like cyber forensic people in order to understand what's going on behind the screen. So first started with like some kind of simple maps and how basically data is flowing, what's going on. What is basically the first question that we asked was like, what is the life of one internet? And then we started to follow stories were coming more and more after that. 
+
+Vladan Joler (00:03:07) - And then, you know, like step by step, we were some kind of, we were like discovering different layers of transparency, being able to go deeper and deeper and deeper behind those infrastructures. And then after that, the more and more complex maps, I was able to create more and more complex maps of those. So the first big one was, what about Facebook cognitive outcomes? This was called Facebook. And that was the beginning of those big black maps that I was like, then make Kinglake for next few years. Like the maps, different kinds of questions that I'm asking. Yeah. So that's kind of a brief intro into what I was doing for the last several years. 
+
+James Parker (00:04:03) - Um, thanks so much. Uh, when you, when you say several years, I mean, how long have you been doing this kind of work? Is it, you know, is it a decade now or, 
+
+Vladan Joler (00:04:12) - Well, I think that the shared lab, I remember I was maybe like, uh, okay. The, the, the same conference and those kinds of big evidence. There were like 2011, let's say. And, but those more kind of like, uh, visualization and visualizing invisible infrastructures, I think it started like around like 2014 or so. 
+
+James Parker (00:04:40) - And would I be right in saying that anatomy of an AI system is the kind of highest profile of those visualizations? 
+
+Vladan Joler (00:04:47) - It somehow became the most visible one and it's a bit different than the ones we were doing before. So in the beginning we were more like interested in, so it's a process of, you know, it's a process of learning what's happening.
+
+Vladan Joler (00:05:04) - So first we go to like being satisfied in a sense, just to getting a picture, you know, seeing the map, seeing the infrastructure, then, you know, more and more, we were like going deeper more and more, the, those maps became like more and more abstract in a sense. And the questions that were asking were different  ...  and AI, it's kind of like different than other ones, because it's not so much about the infrastructure. It's also not so much about like data visualization. It's more like a cognitive map of one really, really complex system. And somehow that might probably resonate better with the general audience. And what was really interesting is that, uh, you know, like there were like so many different audiences that were like accepting this fact as something interesting because in the beginning, like those first Maximus making was mostly like communicated within the tech community and also like, uh, legal and advocacy communities related to this kind of decentralizing, but also like privacy issues. And somehow with the, with the, that, to me, if an AI went into completely kind of different directions, so first people started to realize the, or like accept the system, kind of art design, kind of like a tool for, um, you know, like universities and teachers to explain something. So it went into lots of different directions. So probably this is why it became more known than the other months and went out of these kinds of circles of like internet activists, lawyers. 
+
+Vladan Joler (00:07:08) - And I'm kind of lucky that we managed to, to break those other bubbles and be present on different places 
+
+Joel Stern (00:07:16) - I was going to ask, um, flat and what, and this is, you know, again, I feel a bit silly cause there's sort of some things we talked about last week, but I think it's, it's, it's good to cover that ground in the, in the hope of sort of progressing a bit too, but just to say something about why you chose the Amazon echo is that as the object of study and, and, and what it is about that particular device that, that, um, made you feel like, um, such such a complex map would be possible. Um, but, but more than that, um, you know, sort of productive and generative, 
+
+Vladan Joler (00:07:51) - Well, it was some kind of, you know, there was like lots of different, uh, aspects of that came together in like why we chose Amazon. First of all, were like, uh, involved in some of these sorts of projects, Mozilla foundation about like those kinds of voice interfaces in a way, like, for me, you know, like this was like kind of topic that we were like researching and, and those Amazon echo at that time was one of the most known and one of the most exposed commercial objects that's died. And then I met, uh, the same, like, um, working process. I met Jake and, and then we kind of started like together to explore this idea of visualizing what's going on behind that specific for me in general, it was maybe not so important is it like Amazon echo or any other device, but somehow like, because it belongs to, because at the end, like those complexities are similar to many different options. 
+
+Vladan Joler (00:09:09) - So in a sense, like if we think about anathema and AI, the map will not be much more different if we speak about like an iPhone or any other CD, for example, or something like this. But in a sense, I think it was good to choose, uh, Amazon product because Amazon it's like one gigantic company and, you know, like going through those kinds of jungles of Amazon company, there's a lot of interesting things to discover, you know, like starting from this kind of like distribution centers treat their workers and those kinds of automatization of labor. So it is a really, really, and in the sense, like, you know, develop this man is owning this in the world is owning this company. So like, it's a good example to explore this complexity of contemporary capitalism, contemporary.
+
+Vladan Joler (00:10:15) - No production chains. Yeah. I think we, we, it was a good choice compared to some other device. 
+
+Sean Dockray (00:10:27) - Can you talk a little bit about how you went about doing the research or like what information went into the material to make them? 
+
+Vladan Joler (00:10:37) - Yeah. Yeah. So the, the, the, the middle part of it, I could open the map. 
+
+Sean Dockray (00:10:41) - I was more just thinking, you know, where you look like, what a database is, what I'm. 
+
+Vladan Joler (00:10:51) - Hmm. So, Sarah, there are many, many, you know, for me, uh, to those segments of the bank can be in a way it's a, it's a story for itself. So there is no one way of trying to deal with all of those things, because like, for example, if we start with, with, and this is in a way similar to how we, how I progressed in my research, you know, it always starts with you and then trying to go deeper and deeper to those like different layers of transparency. So I spent like, let's say first five years in, in doing this, doing this middle box. And then when I started to throw the, the, the, the anatomy of an AI, I knew all of that in a sense. So that was not the problem. The middle part of the map was never in that moment. It was not the problem for me. 
+
+Vladan Joler (00:11:57) - So, so for example, just to explain your different kinds of like the first you try to open the bikes, you know, to see what's going on inside, then you'll realize that it's not much to see. You know, then after that, you're trying to understand how these devices connected to my mentor and what kind of life that kids have, what type of data is going out from device and entering into device. Then you're trying to follow those packets to understand how those packets and where they're flying and destination. So this is linked. For example, then you are researching the locations of those like data centers was trying to understand how this in the visible infrastructure of Amazon data centers is working, which data center you belong and your data have belong. And for each of those steps, you are using different tools, different types of like different methodology. 
+
+Vladan Joler (00:13:02) - So for example, for, I dunno, you can try to find the blueprints of device, and then you are entering into some kind of shady websites in which like sharing the blueprints of different objects. And then you realize, okay, this is the first layer of transparency, because like most of those companies doesn't want you to be able to have a blueprint of their conflict in a sense of not just that this is some kind of like a business secret or something like they don't want you to be able to repair devices. Well, you know, most of the cases, for example, in case of Apple and stuff, then I'm spending a lot of time to try to find the blueprint, for example, to be able to make some kind of drawing of that, then next step it's like, for example, like  ... , you're using the, and then you're trying to understand, like dropped her off the back ends and what kind of data is fine. 
+
+Vladan Joler (00:14:12) - So those are in the middle of middle of this might, it's mostly like a tech research. So we are using different technical tools to explore. But then in the, in the moment we, this metrics pendants on both on the left and right side. And this is basically with the one friend of mine and, uh, John really, you know, uh, infected me with this idea of like materiality of infrastructure. And then she gave me the book from about geology of media. And then I was completely impressed by this idea of, of being able to look all of those infrastructures, but in some kind of deep time from deep dive view. And this is when we expanded this mapping to some kind of light.
+
+Vladan Joler (00:15:13) - Stop geological process elements. And then when, when I, when I stopped that, so it's a different end research that it's a different methodology, completely different methodology, that middle part of the map. So in, in this case, it's mostly like, you know, trying to feed somebody, searched from other people, trying to collect enough information about suppliers. So it's mostly, let's say investigative journalism on one side, but also some kind of like, because I had the chance to more or less piggyback on this research from what was the line in the sense of like, I was going to go around the world and doing somebody to IOT. And then I was running around, around in the way, because we were going to India, China places where basically materiality of this kind of production is happening. So I was running away and trying to do my own investigation, trying to visit mining sites, go mining, or trying to Foxconn in China, trying to, not trying to see all of those places. 
+
+Vladan Joler (00:16:40) - And for me, that was like the really, really important parts. Even most of those things you cannot see on this map, either you read in the text at the end, that's kind of expedience. So they're being, they're trying to enter into some building complex. So it was really important for me to be able to feel the thing really going on. But in a sense, you know, like most of the left part of the map is mostly like, uh, let's say investigative journalism and maybe some kind of academic working late, fine to collect as much information. 
+
+James Parker (00:17:21) - Can I ask Vladan um, did you get into Foxconn or any of these places and, and what was it like? What did you learn specifically? You know, you said it was very important. What was the, what, what was it that was important? 
+
+Vladan Joler (00:17:43) - We, we, I didn't manage to Foxconn. I was in front of the building and then looking out those like workers and stuff like you, I didn't have enough connections or whatever, get inside the folk school, but we managed to go into one of the biggest Amazon houses where those robots are going to cut down and picking stuff. I managed to went to a few, uh, mining sites and I managed to go to one really amazing place where for example, all of those ships are dying. So it's some kind of graveyard of ships in India, but not in, not into Foxconn at the end, but, um, yeah, maybe next time, like, because for example, in case of Amazon warehouse, I think it's important to understand it's different when you're looking at this from some pictures or looking at this place from like, by reading some patterns it's different when you're, when you aren't there. 
+
+Vladan Joler (00:19:03) - Uh, you know, physically, because for example, in the, I was never imagining, for example, this kind of, almost of the big house could be so noisy so that the level of noise it's like amazing and all of those people there. And I'd say it's like really, really feels you're in some kind of inside the body of some Machine. And I mentioned to here, I forgot to tell him like the last question, the question before, one of my favorite methods of investigation, it's also like digging through different patterns. And this is in a way we did like a big bar to the Facebook map, but the old, uh, three was also like really present doing this one.
+
+Vladan Joler (00:20:00) - What was it like really also like really important for me, in a sense it's like, okay, you get there enough data. You'll know how to parts of this puzzle, but in a way, like really important it's cow at the end, you are shaping all of these into something that is visual. And in the sense like this is like for me, what was really helpful in school, Daisy can try to think about these map through those, uh, triangles that were really present. And they are basically the main, uh, let's say visual tool for, for, for, for this map and this something that I kind of as a, as a, from one book also from pretty standard books about, uh, uh, digital democracies and in a sense, those triangles helped me to, to build this story. So, so basically like, uh, the main idea behind those triangles. 
+
+Vladan Joler (00:21:11) - So you always have some kind of means of production, so resources and tools, and you have a labor and with two of them, you are getting a product of the Lake. And in a way, it is how I shaped them in the sense of like some kind of like continuation of those triangles. For example, you're starting with, uh, you know, some kind of, uh, elements, and then you will, you are, you are mining those like, uh, metals or things, and these four become some kind of resource. And again, using some labor sense of smell of thing. And basically following this map, this map, you're following some story that is held to those triangles. That was kind of one of the main core visual elements. And let's just be as relevant logical arguments with these maps. 
+
+Joel Stern (00:22:18) - And w what happens to those triangles when, when we sort of move into the central spine of the map, they start to become, you know, sort of much more complex in a way. And there's one diagram that my, my eye is always drawn to, which I think is the, um, Amazon voice services diagram sort of debt down towards the bottom there, which is that sort of an amazing image of hundreds of intersecting points and on a, on a sort of node that the AVS image, um, uh, sort of need, if you could sort of, um, say something about what that image is, um, sort of representing to you and, and sort of how you kind of produced it from the material. So basically this is the moment 
+
+Vladan Joler (00:23:04) - In this doctor when we were getting into the AI and machine learning, uh, level of the story. So basically this picture is some kind of like, let's say the traditional way, how to represent a machine learning system. And so we took those notes and some kind of, you know, this is like the story about the, either part of the, the part of the system. And he says, yes, the kind of structure it's different in the middleman, or part of the life can drive out because left and right. They're mostly presented through some kind of different way of storytelling. And in the sense that I don't know the most complex part of this, those two drawings on the bottom, and I really spent weeks or months, and like trying to deal with that, because I know from, from here, we are seeing some kind of, you know, the left side of the map. We were seeing some kind of process of flotation of nature and exploitation of human labor. But here from the bottom, we are seeing some process of exploitation of, of date quantification of nature. And then I had this kind of idea of how to, you know, let's try to classify what kind of like the ties being quantified in a sense of like first I.
+
+Vladan Joler (00:25:00) - Uh, tools to, to separate two different fields like quantification of human body actions and behavior, and then quantification of human made products, because idea was like, you know, everything that can be quantified real big because like this kind of, this is this kind of offline data education, everything that, that data is being understood in the sense of like our body and what we can do. And they becomes this territory that need to be invaded. And in a sense here, I tried to classify all of those different, different types of like, so we have like, for example, quantification of our in individual bodies or social buckets, then we have human body biometrics, medical, forensic, psychological behavioral profiling.  ...  
+
+Vladan Joler (00:25:59) - lots of different things because like, when you throw this map, this kind of thing, it's about CPGs. And this is really hard because like, when you have, you know, like, uh, three lines coming from something, you know, this is a statement, this is like your own classification system, you know, all in a way when you have like this human being and four lines, this is how you define something. This is your claim. It's, you're saying like there's four types of law. And in that sense, like it's really hard to struggle or all the time with these kinds of like, you know, like every line leads to some kind of statement of something. And then I spend a lot of time in, in like trying to understand how different people present or in classified something, you know, in order to, in order to be more and more precise in how I'm drawing something, how many lines that are coming from each of those points. 
+
+James Parker (00:27:14) - I absolutely love those, um, those spirals of the bottom of the diagram as well. They're really, really provocative. I mean, apart from just being beautiful on the, on, on the, at the eye as well, and they sort of invite you to get in and sort of turn your head slightly and sort of read, you know, to read them, they sort of invite a slightly different relationship with the diagram, but they're also, you know, they're, they're the bit where the data extraction happens. And as, as you're saying, that's, you know, that's crucial that I was just wondering if you could say a little bit more about how we should think about data in relation to extractivism or, you know, a lot of people saying, uh, you know, um, data is data, production is a form of labor. We need to valorize data production and we can produce a kind of a relationship with the capitalists and so on where we get paid for our data reduction. 
+
+James Parker (00:28:08) - And some people say, no, no, no, no, no. Um, data extraction, you know, is really like a form of colonialism. And we can understand this kind of expropriation, eh, as part of a continuous history of, um, capital's relationship to colonization. And, and then there's, you know, other people saying, well, no, the problem with data is bias or computational empiricism because it imagines that to be a kind of, uh, you know, a truth to nature that only the algorithm, you know, can reveal, um, through the process of data extraction and so on and so on. And I was just wondering, you know, could you, could you sort of say a little bit more about how precisely you understand the, you know, the, what it is that's going on with his extraction of data? I mean, either in relation to this diagram specifically, or more generally, how does data become a political problem for you? 
+
+Vladan Joler (00:29:06) - Yeah. Okay. I think, I think that's the key question then the key problem, the another one it's exploitation of labor and resources, but this question about this kind of new types of activism that exists, they're like really important to try to be on my, I have all the maps it's still being published, but this one is crazy. So I tried to make, uh, new methods called a new extractivism and try to basically this one is a bit different than the other ones.
+
+Vladan Joler (00:30:07) - Because this one is kind of, uh, it's, I'm kind of like mosaic or assemblers, assemblage of allegories. So what I try to do here, I tried to mix all different, like lots of different ideas that I really like on this topic of new extractivism and exploitation of data into one big, crazy map. And I considered the form. These basically starting with this idea of also from one friend talk is some names through the sea, like speaks about his gravity of, of different platforms. Like for example, Google and Facebook, like, you know, like in the similar video of like, uh, this kind of theory of relativity, like how they're like bending the space and time. And in a sense, I, um, he is talking with this allegory and then like, you have like little God. So it says some kind of a big crazy diagram and it's called the new extractivism and it's, it's basically some kind of mix of different, uh, uh, allegories and some different concepts that are combined into, into, into one story. 
+
+Vladan Joler (00:31:26) - And in the sense, it starting with this kind of idea of, uh, starting with, uh, this idea of gravity, how different like, uh, how basically those companies are bending the space and time. So we have like one little guy who's trying to escape from this, like a black hole that we mentioned that a black hole, for example, in one moment you cross this like point of no return and you are becoming this kind of person who is like, you know, addicted to those like services or personal, who is not able to go out from there anymore. And then from here, this person is falling into some kind of another structure, some kind of combination between the cave from a plateau and of course, an optical. So this person is falling into the cave. Then I'm trying to understand what is the texture of this cave. 
+
+Vladan Joler (00:32:33) - And then so on and on, and following like this, basically the structure going deeper and deeper into this kind of really abstract ideas and obstacles, uh, concepts that basically define, or how we live today. So for example, here, for example, you have this idea of, of like you ask about data so that it creates a goal in a sense like that. And now you have like a different, uh, companies, different corporations that are trying to, you know, conquer this and that in a sense it's like this kind of wild, wild West racing their flags into some kind of data fields, but in a way, most of those data fields are related to our bodies. So it's some kind of contest to conquer deeper and deeper and deeper into our body. And in a sense, what's really interesting in, in, in context of like, you know, development of these things, it's like before you had like more or less kind of like, uh,  ...  earth, let's say like how many different metals, how many oil you're able to extract let's in the sense, like in order to progress this kind of like, uh, uh, 21st century capitalism needed to find a new tenant to conquer. 
+
+Vladan Joler (00:34:13) - And in a sense like the, the, those, like  ...  data, it's this kind of new territory that is able to be, to go like, you know, have this kind of affinity to infinity. So there's like so many different worlds that can be almost, uh, infinite, uh, you know, like some number of worlds that can be, uh, conquered, but in the sensitivity, it's important to understand that those, you know, girls are basically our bodies and our socials and everything about us. And in a way then in, after you, in a moment, they are able to conquer. Then there is this.
+
+Vladan Joler (00:35:02) - Second step in the process, it's basically the process of them closer. So they are trying to prioritize and to build the fence around those territories. And once they build a fence, they're able to gain profit and to do the online space based on online disposition. So the, the ECU have these kind of new extract to be some it's like really, really, I think really important, really deep. And in a sense, I also, in some sense, follow this idea of like, you know, like continuation on this kind of whole colonial practices, but the kind of like a lot of the demands. So it's not just about like data is a new field it's combination as if we combine this idea with the things that you're like explaining. And then that, to me, if an AI, then you'll see that all the time you have this kind of mix of different forms of exploitation, human labor, and exploitation of nature. And that, and I think it's like really important to, to speak about that in that context of like, not to have like, you know, to have a cold picture, we need to see the problem more broad in the sense of like covering social relations and labor and stuff. So that's the, but this one it's maybe too crazy for like, maybe we can speak once, like, like in separate topic about this one. 
+
+Joel Stern (00:36:50) - I love this one. Vladan I mean, this is amazing to me. I mean, it sort of goes so far beyond the kind of material infrastructure that you were mapping before. Is this still a sort of work in progress or is this something, 
+
+Vladan Joler (00:37:08) - Hey, you totally need to proofread it too. Like I was like speaking with like friends and colleagues that can comments and stuff. That'd be, so it's almost done. 
+
+James Parker (00:37:25) - I was wondering, I don't know if it's too early to ask this question, um, but it's sort of clear that so much of your work is about a kind of, um, I want to say, you know, a diagnostics or a representation, right. And sort of, there's a sort of, there's a politics to that representation. Obviously, you know, we see the world differently. I, I see the world differently. Having seen anatomy of an air. I can't, I can't get it out of my head. You know, it's, it's a frame through which I now encounter, you know, not just Amazon know Alexa or echo or whatever, but also, you know, now the halo, which, you know, just got launched last week and wants to listen to our emotions apparently, you know, and everything. So it becomes a frame. So there's something obviously strongly political about that, you know, this kind of map that's working at the level of, you know, a little bit more speculatively or something too, but I wonder do you also put on a slightly more expressly, normative hat, um, ever, do you ever think, you know, the, sort of the, what is to be done question because there's a sense in which like the, the maps are sort of, so they, they point to, you know, the whole world and the G you know, the, th there's something overwhelming about them as well. 
+
+James Parker (00:38:48) - So there's, you know, one might one might encounter anatomy of an AI system and, and feel kind of politically disabled, you know, having had one's eyes opened, or what have you, I'm just wondering, do you ever sort of, do you ever step into that sort of slightly more kind of advocacy space? Do you have, do you have a sort of a sensibility about how to confront the, the challenge of whatever you might call it? So, you know, probably not surveillance capitalism, I'm, don't know what you'd want to call it, but you know, the beast, the behemoth that, um, your you're representing in your diagram. 
+
+Vladan Joler (00:39:22) - Well, I, I think like, uh, you're not like I don't feel, I think my mental capacity in a sense of, like, it's kind of limited into like, uh, trying to understand reality. So I, I don't feel confident in like, proposing 
+
+James Parker (00:39:45) - That's quite a big job on itself. It's not that limited. 
+
+Vladan Joler (00:39:51) - Yeah. Because if, for example, I not, you know, I, for example, I'm not too sure should, might keep.
+
+Vladan Joler (00:39:59) - Go to school tomorrow or not. And in that sense, I don't feel comfortable of like proposing how to solve the world's problem. I'm not able to deal with everyday questions. So what time may be good, just trying to understand how reality works from my point of view, and then maybe there will be someone else who will, based on that came to you. No idea. Okay. Maybe we can do it differently now, but I always like to quote like alone, some people are all asking this question for me, but I, 
+
+James Parker (00:40:46) - No, I hate that question when people ask it to me. Uh, yeah, 
+
+Vladan Joler (00:40:50) - But then the cool thing, this, uh, you know, a guy from blade runner, my job is still to find what's wrong, you know, to deal with the, not to propose how things should be done now. So I kind of, for me, it's enough. 
+
+Joel Stern (00:41:08) - It was that was that the blade runner or, or the replicant who said that? I can't remember. 
+
+Vladan Joler (00:41:14) - I think it's a blade that having some form, the, the, the thing is the problem it's like so complex and so diverse that in, in, you know, you know, for example, there's like this kind of thingy about like, you know, ethical AI and stuff like this. And then the question is like, enrich segment. Do you think AI can be ethical? Are we speaking on the level of data set or we are speaking about level of, of like, you know, like workers in Congress, so, and in order to solve the problem deeper, then you need to deal with much bigger problems, like issue of capitalism, issue of like inequality, colonialism, you know, like, so I don't think there is like a simple solution. What we can maybe try to do is to try to improve parts of the map that is not, I don't believe there is like, you know, like one solution that can solve all of these. 
+
+James Parker (00:42:23) - Can I ask, um, similarly speculatively, uh, if you have an intuition about how the pandemic is affecting the, like, if you were to redraw anatomy of an AI, you know, in 2020, do you think that the story would look different or do you think, do you have, have you been sort of following, um, big techs sort of positioning in relation to the pandemic? 
+
+Vladan Joler (00:42:52) - I didn't follow a lot, but what is obviously the most dedicate part of the mappings? Basically those supply chains, because we managed to dependent on like planetary scale systems of production bananas that are coming close to our shops or how tech is being made. So in the sense of, I remember like signing ex-Yugoslavia and during the eighties, because we were like in between two blocks to the Western block and we, uh, we were pushed to, to develop our own, uh, that infrastructure on factory. So we were able to produce, you know, like our own capacity at that time, of course it was not even near to all like, sorry, in a sense, this is like, but it took like decades to able to do that. And I'm not so sure that we, you know, like, uh, because it's really obvious. It's like those like flex supply chains are now collapsing because of, it's not so easy to see like production on different places in the globe. This is something that is going to collapse. And for example, one friend of mine, just five to buy like 10 the sweets. And there is not because there is all like supply week. 
+
+Vladan Joler (00:44:39) - So this is something that continued to be present. We rebuilt, then it will be like some kind of like a moment. Did we realize how fragile and how conflicts are in our supply chains and production chains.
+
+Vladan Joler (00:44:59) - And this is to be a problem, but there is no easy solution. It's not like we can grow our own, you know, computers in the backyard. Like we can grow. That could be interesting, you know, like the complex technology relies on like lots of different materials coming from lots of different parts of the globe. So in a sense, you know, probably if this continue, we will, I'm not so sure that we will be able to produce this kind of sophisticated technology anymore in the sense of like, because like, you know, not each country or each part of the world have all the materials that are needed. Some kind of like, I dunno, it's going to be interesting, but the, in a sense, like, you know, like we should not believe in trumps and like, you know, like these kind of ecology made in USA, it's like, it's almost like even the big, lots of different resources that probably not in this moment to complete the full supply chain within one country. And then on the other side, kind of like, you know, like don't like, uh, what is also interesting field is how all of those, like data collection and surveillance that is bound to be now used for those tracking. 
+
+James Parker (00:46:42) - Yeah. So Sean's question gets to something that I was wondering right at the beginning about the choice of Amazon echo, you know, there's something about, there's something about a voice interface that seems particularly, you know, opaque, we're not a peak, um, immaterial, um, invisible, um, it, it wants to disappear itself and even more profound a way than the computer with the computers that we're talking through now. So I suppose, um, I I'm, I don't want to overrun Sean's question with my own, but yeah, just this idea of the voice, the specificity of the voice in some way, as in relation to visual analysis. 
+
+Vladan Joler (00:47:37) - Yeah, no, but what's really interesting is it's kind of like deconstruction of the interface and visibility of the interface, because before we were able to, you know, see the interface, understand the element and something to face, and then in that sense, you know, becomes some kind of like your office, you know, that can defend and defines what you can do or what you can not do define the length of your, how many characters who can say, or write or what they would. And in a sense, like that sense the visual interfaces are, let's say more transparent to salvage extent in the sense of like, maybe they are more transparent, but in a way what's going on behind them, these kind of invisible infrastructure and the rules them are, again, not so transparent with the voice. We have a similar, even more maybe complex situation because we are not able to see data. 
+
+Vladan Joler (00:48:53) - We are not able to heal the borders of the interface. It's not written, it's never written, right. But this is the least of the works that this system can understand. For example, in case of Amazon Alexa, like those of 200,000 words that this device can understand. So in a way you don't know the borders, the system, I think that you know, where those borders exist. And then in that sense, those borders are one that are defining what is possible and what is not possible and what is possible for, or different people. So for example, for male or female, for like different decks, probably it's much harder to deal with those kind of like can visit completely invisible, completely invisible interfaces. But I believe there are ways to investigate that as well.
+
+Vladan Joler (00:50:01) - Let's see, I think it carries the similar function of like the normally regional interface in a sense of like, it's kind of cool, active. Maybe it's even more regular than the visual. No, no, no. I didn't went into that direction of trying to understand the Porter, trying to understand the shape of the interface into some other direction. We had like one project that they never managed to finish. But then thinking about the law, it talks about the border. It was about like investigating where is the border, for example, limitation, for example, like, uh, in that case it was about Google. And the question was, is there any war that exists that is behind the border? So that then any, any war that is not captured by this Machine, and then trying to understand what will be behind the border, but try and understand that you were like doing like different kinds of tests. 
+
+Vladan Joler (00:51:15) - And then for example, you find out that there are some kind of limitation. For example, I don't know, it cannot be more that is longer than 125 characters. So in the sense like the Google crawler cannot pick something that is bigger than one 25 characters. And that sense, you know, like all of those worlds that are bigger than 125 characters living in some kind of free world without Google, because they're too big to be captured because all of those smaller words are basically transformed into profit capture on the big horse. Well, in a way there is no, I think there is no bigger words than 50 characters that's about, but then like going through through there, I realize, okay, what if we write without a comma space and adopt and write all the, you know, like things together without separation between then? And I tried. 
+
+Vladan Joler (00:52:21) - And in a sense, like for example, there is no, you should find some kind of longer sentence without spaces. Google will not be able to recognize this. So it's really important to understand where the border is because in the moment we understand what is the border of the capital interfaces, similar thing, then we are able to play with it. We are able to explore what's going on behind and to explore maybe how to protect the species that leaves behind the border, some kind of free zones and stuff like this. So this is what I'm really interested in. And this is why it's really important to try to understand, to understand what are the limitations of those systems, including the Moxy interface. 
+
+Joel Stern (00:53:18) - Yeah, I, I think, um, that that's, that's an amazing sort of thought to, to sort of, um, con to conclude with, because that's something we've been, we've been thinking about and has come up in every, every conversation we've had really, uh, I think around, um, questions of intelligibility and the end, the possibility of escape and the possibility of, you know, um, thresholds, um, beyond which the logic of these machines and devices and sort of extractive systems, um, CA CA can't function. And, and, um, we've been thinking about it a lot in, in, in relation to Machine Listening specifically, and, you know, one of the, um, names we've given to that sort of line of thought is the lessons in how not to be heard, which, um, some of the different people who sort of participating in the project have interpreted sort of as tactics or, or strategies for kind of communicating in ways that say, you know, Listening machines would, which would be unintelligible. So it kind of completely analogous to the kind of, um, practices that would be answered and sort of searchable, but by a Google engine, you know, at the same time have had other people kind of, um, come back and say, well, some of those individual practices of evasion are a bit sort of, um, you know, they're sort of not structurally or politically kind of, um, impactful enough that they're just sort of, you know, ways of creatively.
+
+Joel Stern (00:55:03) - Evading capture temp temporarily or, or training the system to, you know, um, decode ever more complex material, but in any case, I think it's, um, what I really, they took from me from the other document, which, which you showed, you, showed us the new extractivism, which is really amazing. I'm really sort of thankful to see that work in progress with these, uh, that there is a desire to escape it, however impossible or sort of temporary, that escape might be, you know, the black hole is there, but the little guys like tried and still trying to get out of it. We should let you go. Um, I think it's, it's been a great conversation.

content/transcript/maier.md

@@ -0,0 +1,79 @@
+---
+title: "Stefan Maier"
+status: "Auto-transcribed by reduct.video"
+---
+James Parker (00:00:00) - Well, Stefan would you maybe just begin by introducing yourself, however makes sense for you.  
+
+Stefan Maier (00:00:06) - Yeah, sure. Yeah. So, um, yeah, I'm Stefan Maier. I'm, uh, I'm a composer, uh, primarily of electronic music. I'm, uh, born and raised in Vancouver, Canada, which is where I'm I am right now. I was in Berlin for a little bit working as a composer, and then I studied in the States and now I teach at Simon Fraser university here in Vancouver. I teach electronic music. I kind of work between the experimental electronic music world, um, contemporary classical music, and then multimedia installation. What kind of unifies? All of it is a examination of both emergent and historical sound technologies. And in that I try to highlight kind of material instability or unruliness. And, um, yeah, I'm really interested in kind of mapping the flows of kind of chaotic Sonic matter. And in that mapping, I'm always trying to uncover alternate modes of both authorship and listening practices possible within this specific technologically mediated situation.  
+
+Stefan Maier (00:01:15) - Yeah, I guess for me, like really one of the primary departure points for my work is this idea of the prepared instrument I'm trained as a classical composer primarily. And now all of my work is always kind of trying to investigate the specificity of the, um, of the, the technical apparatus itself and then trying to coax out, uh, a certain kind of logic, a certain kind of strangeness and attending to the specificity of that. So the reason why I bring up prepared instruments is just because that's kind of been an ongoing fixation of mine. So, you know, cage John cage puts bolts and various objects into piano strings and do familiarize. Is this kind of like historically rot instruments such that, you know, it creates this kind of a sound, which, you know, he was aspiring to make it sound like a gamble on an orchestra. 
+
+Stefan Maier (00:02:08) - And there's a long history of preparing instruments, kind of tinkering with the instruments and dealing with like the technical possibilities of an instrument to bring out kind of like a, let's say, repressed character of that instrument or transforming it into something else. And I guess for me, that's kind of like a primary point of departure regardless of what I'm I'm doing. And like I say, that manifests itself in a variety of different contexts. Like I have a classical music practice, like I, I write for ensembles, uh, quite regularly. And then I, yeah, I, I do a lot of improvised electronics and then I use kind of increasingly larger scale and multimedia installations. 
+
+Joel Stern (00:02:50) - Um, Stephan, could, could you just expand a little bit on sort of the idea of, you know, Machine Listening at, as a kind of pre prepared instrument, you know, Machine Listening is as something that can be instrumentalized by a composer? 
+
+Stefan Maier (00:03:05) - Yeah, sure. Happily, I guess, yeah, much like, I mean, in parallel to the, the idea that I was talking about with the piano I'm with, with these Machine Listening tools, I'm really interested in kind of determining the technical possibility space. That's kind of given to me and trying to place these tools in context where the kind of like latent unruliness can kind of come out. So for me, oftentimes I'm using these, uh, these technical objects as kind of black boxes, which then I place into specific contents that are really outside of where they were designed for. Um, so using speech synthesizers, but not speaking to them, uh, you know, in the way that they were designed, but rather, you know, having them converse with yeah. I don't know, like a soundscape even, or yeah, just, I mean, in the case of the, um, what I speak about in the dossier, um, this idea of just kind of letting the speech that the size, or just kind of generate its bosses. 
+
+Stefan Maier (00:04:06) - I mean, sometimes I'm also actually intervening with the code. So I work with technologists who, uh, well, possibly mistrained, um, a neural network. That's what I did in, uh, in a recent project is DB and chain project where we mistrained, uh, uh, speech synthesizer such that it generated its own language, um, based on it's kind of incomplete training. And for me, I really see that as being parallel to this logic of taking kind of a, ready-made a cello, uh, a piano or something, and then trying to, um, yeah, remove it from its context, such that then it starts to do something else. But again, always specific to the technical possibilities that are forwarded. Like I'm not really interested in kind of like, like fanciful alteration. I'm interested in actually investigating what's going on under the hood. If that makes sense.
+
+Sean Dockray (00:04:58) - I think thinking about under the hood and speech synthesizers, you, your curatorial essay begins with a discussion of Wavenet. Uh, and I was just wondering if you could, uh, talk a little bit about what it is about Wavenet that sort of interests you and how Wavenet kind of acts as this entry point into, uh, your particular interest in Machine Listening. 
+
+Stefan Maier (00:05:19) - Wavenet was kind of like the, um, the catalyst that got me interested in all this stuff in the first place. I mean, I, I wasn't particularly interested in even, even artificial intelligence up until I read this, this essay that Google released when Wavenet was first kind of, uh, released, I mean, on one hand you have a tool which is, you know, one of the most streamlined kind of speech synthesizers ever, ever made. You know, as I know in the essay, it kind of easily passes the Turing test. It's extremely mutable. The training is kind of like, you know, it has no parallel and it's like the technical achievement in so far as it can, you know, I mean, it's now employed in Google assistant and all these things, but I was really interested that even in this kind of, um, as I say, kind of like streamlined application of Machine Listening, you also have this possible, this possibility of kind of, I guess what I S I S I think about it in terms of like, really like digital objection or something like this, where, um, when the speech synthesizer is left to its own, you know, L left on its own or it's, or it's allow to kind of just speak freely without a human interlocutor, you know, it generates these strange glosses, which correspond to, to know kind of like known human language. 
+
+Stefan Maier (00:06:41) - So I was really interested in this kind of this duality, I guess, that here we have, like, Google this kind of technical giant generating this tool, which is like, kind of incredible to, uh, if, if you're trying to, yeah. I mean, have, um, computer human interactions be as seamless as possible from kind of a normative perspective, but then at the same time with the same code you have, um, this kind of this strangeness, which can kind of come up to the surface. And so I was really interested in kind of like parsing out how, how to reconcile that ambivalence, that, that parody, I guess the, yeah, the, the technology's oscillation between like, you know, it's, um, technical constitution, which is capable of both realism and objection, and then also what we've projected onto it as being this kind of the voice of Google assistant. And so, yeah, I guess for me in the dossier that that's a, that's a really central paradox, the oscillation between, uh, rational objectivity and then a projected subject position, which is kind of never, never really fully congruous, if that makes sense. So, so wave not really seemed like a, yeah, a really good starting point for speaking about, um, yeah, a lot of the artists that I was interested in curating, especially somebody like, uh, George Lewis whose work, I think really, um, deals with a lot of these, these ideas of at least implicitly 
+
+James Parker (00:08:12) - I'd love for you to draw out that connection between Machine Listening and music or a little or composition, or a little bit further, because, you know, it really comes through in the dossier that a lot of artists who've been working musicians, who've been working with Machine Listening and for a very long time. And as far as I can tell, you know, a relatively early stage in the project, it's, it's really a musical context that the phrase Machine Listening starts to be taken up and used regularly as a result. I think of Robert Rowe's, um, book, uh, interactive music systems, um, in the nineties. But, but you know, you're, you're also that you, that shouldn't get sort of bogged down in the, in the word or the phrase, because obviously, um, somebody like George Lewis is working much earlier than that, and you discuss other artists too. Could, could you say a little bit about the, kind of the history of Machine Listening techniques in music and the relationship between those two fields, to the extent that there are different fields at all, and, you know, also like the methods, because what you were describing in terms of determine more is not on my rudimentary understanding of George Lewis.  
+
+James Parker (00:09:20) - Exactly what he's trying to do with the Machine Listening systems he's working with. So, yeah, that's a very open-ended question, but I just wonder. Yeah. 
+
+Stefan Maier (00:09:28) - Yeah, it is. And it's, and it's kind of difficult to speak to in some ways, because like,
+
+Stefan Maier (00:09:33) - I'm thinking about Machine Listening, both in the context of applied artificial intelligence, but then also in a broader way. And as you know, like George Lewis, do I think that he anticipates a lot of the things that I I'm interested in that applies specifically to, um, AI stuff. I mean, he, he hasn't really worked with like neural networks at all. I mean, that, that was technology that wasn't really, you know, at his disposal at the time he's using, he's using Machine Listening and more like, kind of a, a broader sense. I'll, I'll speak a little bit to the, to the history of yeah. The uses of specific, uh, yeah, kind of, I guess you're right with RO it's, it's more about like interactivity is really the crux of the matter. And I think you're right. That, I mean, I sent that email. I, I kind of, um, yeah, I looked through some of my resources and it seems like that really is one of the earliest examples of that. 
+
+Stefan Maier (00:10:29) - Um, so it, so it's a fairly recent phenomenon, but it's a very old, there are, there are other things that are in the history of electronic music with the chart anticipate a lot of these ideas that, um, yeah, I mean, go back even to the, to the sixties, I think about somebody like Xenakis where, uh, Yanis Xenakis, the Greek composer and his use of these like kind of highly formalized mathematical systems that are influenced by kind of ideas of human listening and then kind of, um, allowing those, uh, algorithmic systems to basically generate these extremely kind of incomprehensible jarring compositions. And this is, you know, this is, you can also speak about somebody like David tutor who was doing something similar or Subotnick in terms of really like the first person who's using Machine Listening software. I mean, I would say that it's, it is George Lewis where it's like a live interactive system where you have kind of Machine, which is, um, actively responding to input, which is coming out and then changing its behavior based on that.  
+
+Stefan Maier (00:11:35) - I mean, after rainbow family, which is where the first software was developed, he created a system called Voyager, which I believe does start to use kind of more, um, at least its most current iteration has some Machine Listening underlying it. Yeah. So I would say that he's kind of like a major kind of, he anticipates a lot of the things that we're seeing now with like, you know, Holly harmed in and, um, an empty, sad and uh, Jennifer Walsh and all these people. So, so I guess, I guess my answer is that George Louis really is this kind of like central figure. It's hard to generalize about, you know, what's actually going on like under the hood because the technology has changed so much and frankly I'm not a computer scientist or even a music technologist to release speed to that history in any with really any depth, I guess, um, 
+
+James Parker (00:12:27) - You know, placing Machine Listening into a history of sort of proto Machine Listening is really an important thing to do, uh, at the same time as retaining some kind of specificity. I mean, could you maybe talk through some of the different artists works that you gather in the dossier and some of the different things they're doing with Machine Listening and all proton Machine Listening. So you've already mentioned a rainbow family by George Lewis, but is there, is there a sort of another good, good entry point into the dossier for you? 
+
+Stefan Maier (00:12:56) - Yeah, sure. I think that, um, Jennifer Walsh's, um, entry, which is ultra chunk, which, um, she worked, uh, with the technologists artists memo Atkins, who's kind of like a deep learning guru guy. I think that, uh, like Jenny's contribution is kind of yeah, in a way. I mean, uh, her and Florian are the only artists in the dossier, Florian Hecker, the only artists in the dossier who were working specifically with kind of like deep learning, um, Machine listen software, her work is really fascinating. Um, in my mind, especially in light of like this, this, this, this conversation of like a functional utility and then kind of objection where she improvised for many, many, many months and, um, kind of cataloged these improvisations and then Machine listener was basically trained on all of her improvisations. So it was this kind of like this, uh, musical subconscious or something like this, which then she improvises with in real time.  
+
+Stefan Maier (00:13:57) - I mean, it's also like an extremely bizarre piece of music. Like it's just like so strange. I like when I, when she sent me the recording, I was just like completely floored because it was just like, so compositionally bizarre. And she speaks very freely about like this premiere at Somerset house in London where, you know, she was just kind of like blown away by, you know, how she both identified certain elements of her compositional language to what was coming out. But at the same time she felt like, you know, there were certain issues of like timing and also of kind of let's say improvisitory syntax, which were presented in this like totally abstracting and kind of warped melted way where, um, yeah, it took this kind of imperfect mirror where you think that you're projecting this, um, self portrait, right. This very like intimate thing where she's improvising with herself every day and this like kind of daily practice almost. I mean, I know my improvisitory practices almost a meditative practice and then having this very personal thing kind of.
+
+Stefan Maier (00:15:01) - Exploded and, um, transformed into something which is deeply uncanny and unsettling and both for the, for the, for the performer. And then also for the audience member. I thought that, yeah, that was really kind of striking. That would be this kind of, this, this is intimate gesture of kind of offering something to the, to the black box and then having something totally, you know, abject come out of it. Then on the other hand, you have somebody like Florian Hecker, who's working with these very specific Machine Listening algorithms, which are designed to imitate the way that the human ear Prestos as Tambor like the quality of the sound. Yeah. It's just like totally, you know, kind of cutting edge, um, computational model of how the ear presses this kind of, I guess, parameter, which is, um, very nebulous. Like if you read any literature on, psychoacoustics like the science of how humans kind of process, um, sound and how human, the human, uh, human society, um, actually, um, like really parses audio, um, Tambor is a very nebulous category that oftentimes described as like a wastebasket category. 
+
+Stefan Maier (00:16:10) - Anyway. So it's just kind of this algorithm, which was, um, kind of designed to, to unpack tambour, at least the way that it works cognitively and Florian basically put in, he had the, uh, uh, previously written composition, basically resynthesize by this, this Machine Listening algorithm and what ends up coming out is, again, this like very worked strange distorted thing, which doesn't draw a comparison to the, to the original. So it's just kind of very detached, let's say process where there is a rational, um, kind of scientific model of listening, which then also distorts Florian's, um, already very, very formalistic kind of compositional kind of practice. Um, so I feel like those are, those are two kind of like very radically different approaches that kind of like come back to this kind of this, the abject unruly output of, of these specific tools. Yeah. And then, and then on the other hand then you kind of have Machine Listening, being dealt with in a more, um, a broader, a more poetic way, I think with, um, with both Ben Vida and C Spencer Yeh where Ben basically, um, had a, a text to speech synthesizer, um, reading some of his kind of concrete poetry.  
+
+Stefan Maier (00:17:29) - And then from that, uh, and, and oftentimes, um, a lot of text to speech synthesis is now employing kind of deep learning. So as to have a more realistic model of the prosody and also the pronunciation of certain, um, phonemes, um, and he really worked with the kind of the idiosyncrasy facilitated by the specific text to speech synthesizer, and then that kind of produced all these interesting rhythms and yeah, it's kind of an in dialogue with much of his previous work using a similar process of translation Spencer on the other hand. Um, yeah, I was really interested in using also, um, uh, Texas speech synthesizer, but to different ends. Um, he took three different models of three different kinds of, let's say, uh, yeah. Dialects of, um, Cantonese, I believe. Yeah. Is Cantonese, which, uh, yeah, also the model was based in a Machine Listening and then he fed the same kind of the same tech next to the, to the three different speakers, um, such that, you know, like the Texas speech synthesizer became kind of like confused and it created all these kinds of also strange kind of sounds. And then Spencer kind of internalized those sounds and started to imitate them as kind of a fourth, third or fourth voice, which is kind of in contrapuntal dialogue with those things.
+
+Stefan Maier (00:18:50) - So you have this kind of, this, this feedback network, let's say between the technical distortion and then, um, Spencer imitating that for me really also, uh, you know, a really important, uh, person that, you know, couldn't contribute to the dialogue of the dossier, but rather looms kind of above it in many different ways as the American composer, Maryanne Amacher , who was throughout her entire career was very much interested in the idea of computation, uh, assisted listening. So using like different, she was, she was kind of an, uh, uh, devoted to people like Ballard and like the JG Ballard and like status law lamb. And she was kind of interested in different futures where humans would be able to use different programs, such that they would be able to hear like as a different animal. So have, uh, a program which can make you hear as a whale or can make you hear Beethoven underwater, or hear a Beethoven.  
+
+Stefan Maier (00:19:48) - Yeah. The same Beethoven symphony, you know, uh, under the atmospheres of like, you know, kind of like, um, the Cambrian period or something like this. So that, that's something that you wrote about a lot. And then in her, um, unrealized media, opera, intelligent life, she kind of imagines an Institute for computationally assisted listening where you can actually, you know, kind of export the way that any individual's human kind of listens and then have that be a program which could be something that then somebody could, so I could listen to as, as you, I could listen to as whatever Sam cook, I could listen as, uh, early human, whatever. Um, I mean, AMA his writings are all about this and she, yes, she had this massive project that was supposed to be, uh, a kind of a television mini series, which was all about, uh, yeah, very like campy a mini series, um, aware. 
+
+Stefan Maier (00:20:41) - Yeah, there's this Institute is kind of grappling with the epistemological issues of experiencing sound as an other. And so, so Amy Semini at the, um, uh, American musicologists. Yeah. She, she writes a little bit about the context that this work was coming out of. One thing about Marianne is that like, I mean, basically everybody in the, who participated in the dossier is at least deeply influenced by, uh, Marianne or knew her in the case of George, um, Florian's work foreign hackers work is really kind of, I think, a continuation in many ways of, of amateurs project of kind of, yeah. Literally trying to use technology to, to kind of like listen in, in kind of a, a different way. Yeah. I mean, I could speak about the, the, the others as well, if you like, Oh yeah. Terry tablets. Yeah. Derek tablets is, um, presented, um, the liner notes from a record that he did, I guess, in the nineties Terry's work doesn't really, really has never worked with, um, Machine Listening software specifically, but, um, was really, yeah. 
+
+Stefan Maier (00:21:51) - I mean, it's a record that came out on this kind of like glitch label mill plateau in the nineties, and Terry's, um, really fascinated by this idea of specific technologies, having an understanding of, um, a normative human underlying them and then using technology to kind of just store that image. So, um, yeah, basically, um, in a couple of Terry's records, there are a number of different source materials, which are oftentimes very charged in terms of like the gender politics underlying the material that's employed, which are then kind of destroyed or, um, yeah. Made kind of unruly through these different technical operations, like in, in, in tablets is writing, there's almost this idea of like the technology being a form of like, of drag for this, uh, for, for the original source material, or it becomes something else. Um, in terms of, uh, uh, kind of a non-normative gender position, you know, through this kind of technical processing. Yeah. I think that, I think that that pretty much sums it up. 
+
+Joel Stern (00:22:57) - No, that was great. Stephan, I mean, th thank you so much for gut for going through all of the contributions. It's, it's such a rich dossier and, um, you know, the, the contributions are so creative. Um, I mean, it sort of was occurring to me as, as, as you were, as you were describing them one after the other, that sort of, so many of them kind of hinge on this dissonance between the kind of the human and the machine and, or the sort of original and the rate and the reproduction and the sort of human qualities as the sort of exit, you know, I guess the Machine and the human both have a certain sort of excess that gets re reproduced in the work that kind of.
+
+Joel Stern (00:23:40) - The element that's kind of in commands. Sure. It, you know, to both it's re it's sort of does seem like one of the artistic projects around Machine Listening is to sort of continually point to that incommensurable difference, which is often, you know, the project of let's say the, um, companies, the big tech companies producing Machine Listening is possibly to kind of oblique her, right. That difference, or at least to sort of make the human subject somehow indistinguishable, let's say from the voice assistant. And one of the things that we, one of the feedback that we got from unsound on our sort of initial, you know, essay was that it was sort of quiet, you know, pessimistic and critical and sort of focusing on a certain sort of techno politics that's always already captured by capital and is in a way leaves us sort of needing to work against Machine Listening. 
+
+Joel Stern (00:24:37) - You know, so the title of the first zooms sort of session we propose was against the coming world of Listening machines. And I'm just sort of wondering, um, how you feel about those, you know, utopian and dystopian kind of horizons. Um, I think, you know, one of the great things about the works that you described is the imagination is sort of put to work in really positive ways. Um, the quest, these technologies are not taken as sort of, um, intrinsically, um, repressive, but sort of as, as platforms for some kind of, if not emancipatory, at least unruly, you know, as you've put it, um, kind of expression. So I wonder if you could say something about your sense of the politics of Machine Listening across this sort of spectrum from the utopian to the dystopian horizons of it. 
+
+Stefan Maier (00:25:35) - Yeah. I, I think that for me, a really important distinction to be made when speaking about adding technology is the distinction between like a, uh, the utility of the technology. And then it's technical operational logic, like the operational logic, which is underlying it. Um, and that's something that I I'm very much influenced by the thought of jail, Betsy Mondo, um, especially the way that he distinguishes between how a technology is used within a certain cultural context. And then what the machine is actually doing. And Simone doll, um, you know, speaks about, you know, kind of the, the alienation of subjugated peoples to, um, the machines in terms of not really being able to understand how that technology can actually behave or can be used in, in, in ways that are in contrast to utility. I mean, you know, uh, I was being a little bit, I think, provocative, let's say in the, um, the dossier, uh, by not speaking, uh, very much through the insidious uses that you just spoke of in terms of like, kind of, yeah. 
+
+Stefan Maier (00:26:44) - The vested interest of, of technical comp of tech companies and then also governments. Um, I mean, there's a lot of vital work that I know that you guys are in dialogue with. And, um, that, that speaks to that. I suppose, what I was interested in doing was kind of just drawing attention to the fact that much, um, especially in the history of electronic music more generally, you know, there's been a, a long history of, of people who've been on kind of the outskirts of experimental music who have been using technologies that are designed to do a specific thing, especially technologies that are designed by the industrial military complex, especially in the context of like the vocoder, you know, and then kind of discovering something else. And there being this, as, as you said, I like this term of like, like thinking about the excessiveness, both of technical activity of rationality and also of like the design, like the design, um, of a, of a specific tool. 
+
+Stefan Maier (00:27:40) - One can even think of like, you know, it's like, I mean, the, the Dawn of like, you know, how some techno, um, and of course, like acid house, I mean, it really comes from taking a tool that was supposed to be used for like, like dad rock bands to kind of like jam to, and then discovering that there's this entirely, there's a, there's a new world. There's an, both, both in terms of like, you know, um, aesthetics and in terms of like, kind of the sociality that emerged around this specific kind of music. So I guess I'm like.
+
+Stefan Maier (00:28:09) - To go to the question of, uh, or the duality, let's say between like the techno pessimists and the techno files, I'm, I'm totally agnostic. You know, um, all what I'm interested in is trying to attend to the, um, the specificity of the technical object that we're working with and trying to then understand, you know, what's actually going on, even if it's a black box, like in the context of, uh, Machine Listening, uh, algorithm, that's driven by deep learning. I mean, it's literally, it's an incomprehensible space. It's a end dimensional space of statistical inference that many of the people who are training these things, like don't even really understand what's going on under the hood for me, I guess. Um, what's interesting is yeah. Trying to find the specific tools that will then be able to offer, um, yeah. Uh, determined, um, their, their, their, their Machine Listening softwares that are so trained and so functional that, you know, like the only thing that they can do outside of, you know, the thing that they're designed for in the case of Wavenet is make this in comprehensible babble. 
+
+Stefan Maier (00:29:13) - For me, that is a source of at least some sort of potential emancipatory potential. I don't know, but at least there, there are crafts, you know, and for me, a really important theorist in, in, in thinking about all these things is, uh, is Benjamin Bratton and his thought around, um, synthetic sensation and how, yeah, there's, there's a, uh, an accidental kind of, let's say, I mean, the way that I interpret is almost this kind of like machinic alterity, which is present in certain emergent superstructures, um, in digital technology, which actually might project elsewhere, then the kind of Silicon Valley Valley ideologues who are, you know, kind of extracting the hell out of all of us. And for me, that's, that's at least cause for some sort of positivity, but nevertheless, I would say that, um, it's always specific to, you know, the technology that we're, speaking of, it depends how it was trained. 
+
+Stefan Maier (00:30:08) - Like, you know, for me, I mean, in my own work, oftentimes I'm, I'm working with unsupervised learning. So the data sets aren't labeled before such that the Machine listener is kind of inferring the deep statistical kind of knowledge about whatever it's, it's dealing with in the case of, um, deviant chain, this work of mine, it's like, we train this, this, this Corpus on like reading, like readings of like nebulous, Dani, and then also like Luddite texts and all these kinds of like, totally like a smorgasborg of just kind of like, yeah, different theoretical positions around kind of like philosophies of technology and philosophies of synthetic sensation. And then from that Corpus, um, the Machine Listening software kind of made this, um, you know, if you guys are familiar with these terms of like, like feature extractions and stuff like this, like these features, um, that, that, that the machine is hearing that are totally incomprehensible to us. 
+
+Stefan Maier (00:31:01) - Um, but are nevertheless these kind of like high dimensional, um, parameters, uh, which it seeing as being, this is the most crucial information of Stephan reading this Neil Luddite texts, you know, but it doesn't, yeah. I understand any content, you know, it's just kind of, um, inferring kind of yeah. Uh, this, this kind of deep statistical structure that, you know, uh, has nothing has very little to do price. I mean, I can't, I can't really speak to what it actually has to do with the meaning of the text, but I know that the output is extremely strange because these features aren't correlated to, to any categories that we have intuitively. Um, and so when those things are unleashed, then kind of a strangeness kind of unfolds and, and yeah, like I say, this is something that's very much influenced, I think by, uh, by Bratton's, um, conception of Machine X sensation and Machine thought indeed rationality in some ways as being very different. 
+
+Stefan Maier (00:31:57) - Yeah. Then our ideas of rationality. So for me, that's, that's definitely, um, that has political ramifications for sure. I think that it's, um, it's, uh, interesting maybe to bring up the idea of like in humanism versus post humanism, like I'm very much interested in this idea of like taking like a rational system or rational technical system and, um, not seeing like how I can interface it with my body or something like this in this kind of post-human context, but rather push the technology as far as it will go in terms of what it actually does. Um, and then seeing what happens. So if there's this humanist self portrait, which is projected onto the, the technical activity, but then the technical activity goes, goes elsewhere. And I would see that as being kind of like parallel to, um, this kind of this, uh, Neo Copernican sensibility that like maybe rational activity might be, um, facilitating where, um, it projects us elsewhere than we thought if that makes sense. And for me that's yeah, that's, that's intrinsically political, um, thing in so far as it's, um, it's questioning this, uh, the rigidity, the givenness of the human that we started with.I mean, will that be unleashed, um, you know, uh, will the, will the possibility of that be unleashed, um, in terms of the way that these technologies are being developed? Um, probably not by Facebook that's for sure. But, um, like I say, this is one of the reasons why I brought up, um, uh, Google wave as well. It's like, even with this kind of like hegemonic, you know, kind of, um, force behind it, there's still kind of some sort of line of flight, uh, which is possible. 
+
+Joel Stern (00:33:37) - This is a stupid thing to say, but I was just gonna say that, you know, Skynet didn't know what they were producing either. So  ... .

content/transcript/mccarthy.md

@@ -0,0 +1,101 @@
+---
+title: "Lauren Lee McCarthy"
+status: "Auto-transcribed by reduct.video"
+---
+
+James Parker (00:00:00) - All right. Thanks Lauren, for taking the time to speak with us, would you mind just introducing yourself and saying a little bit about who you are and what you do? 
+
+Lauren Lee McCarthy (00:00:08) - Sure. My name is Lauren Lee McCarthy, and I'm an artist. Um, I'm based in Los Angeles. And I guess in my work, I'm really interested in the way that our relationships and our understanding of identity is shifting, you know, a lot to do with the technologies that are increasingly occupying every aspect of our lives. And so I make a lot of projects that are kind of mixing performance, software, and film and other artifacts. 
+
+
+James Parker (00:00:46) - You know, Lauren is the one that we sort of first came across and that really jumped out to us as being relevant to Machine Listening specifically, or kind of the way in which it isn't, but maybe is in some ways Machine Listening. But, but also I know that that's very closely related to someone. So there are, I don't know if you think of them as a pair or if it makes sense to speak about those works together or, 
+
+Lauren Lee McCarthy (00:01:12) - Yeah, I really feel like, uh, I think of someone as another, like Lauren as a kind of starting point for an exploration. So I'll talk about that one first, I was noticing a few years ago, the proliferation of smart devices and AI into the home and just thinking about what that meant in terms of privacy and the kind of boundaries around a very intimate, personal space, like the home, like, it almost feels even more intimate than your pocket in some ways, because you have these different ideas of what happens there. And so I just started thinking a lot about that and interacting with the devices like Alexa and Google home and things like that. And I guess one thing that I sort of realized in that process was that, you know, because whenever I'm working with an it technology, I'm thinking about like, how, how do I personally relate to this? 
+
+Lauren Lee McCarthy (00:02:09) - And, uh, I think a lot of the time we are given new technologies and it's hard to find the metaphor to understand our relationship to it. And so we kind of just accept it. So I was looking for that relationship and I realized what I felt most was that I was kind of jealous of Alexa that I'm like this very shy person. And especially when I'm first meeting someone, um, it's like very hard for me to figure out how to like get through or get to a place of, of more intimacy. And then I was seeing this device where people just like take it in and pop it down in the house and started talking to about talking to it and sharing everything. And so I was kind of like fantasizing about this idea of becoming Alexa and getting to do that in people's homes. 
+
+Lauren Lee McCarthy (00:02:50) - Um, and so then I thought, okay, well maybe I can, the piece is called Lauren because I designed this service that, you know, instead of Alexa, it's, it's that you can get in your home, people sign up on this website. And then I install a series of devices. So cameras, microphones, door locks, lights, and other appliances. And then I leave and I remotely watch over that person, 24 hours a day, sleeping when they sleep and interacting with them. And they basically interact with me as if I'm Alexa, but they know there's a human there. So I have like a digital voice and I'm using text to speech. And, um, not only can I talk to them, but I can also control every aspect of their home. And so the, I guess I set it up as my goal to try and like become, or to try and be better than Alexa by drawing on like my human, my humanity, or my ability to observe. 
+
+Lauren Lee McCarthy (00:03:46) - And so I would also like go a step further than Alexa in terms of just like taking action without them even asking sometimes just like deciding that that's what they need might need or want at that moment. And so typically those performances would last anywhere from a few days to a week and I did them in different people's homes and yeah, so that's the Lauren project and that kind of opened up a, a larger investigation where then I started thinking about different ways to think about this, this kind of relationship. Um, and so someone was a piece where I guess one of the things I was feeling was, you know, whenever I'm dealing with performance, there's like the participant, the performer, and those are different experiences. And so I was interested in sharing the experience I had as to the performer, cause it was just so effecting. 
+
+Lauren Lee McCarthy (00:04:36) - Um, and seeing if I could bring other people into that. And some, someone was kind of like a larger scale. So there were four different homes at the same time that I'll have this system. And then you go into a, a gallery and there's kind of, it looks kind of like a command center. It's like some kind of cross between something like we work center and something and like a call center. And you're sitting there with headphones and you've got a laptop, there's four different laptops and they go each one points to a different home. And so the visitors could come and actually just like.
+
+Lauren Lee McCarthy (00:05:07) - Peek into people's homes and then control them and fulfill that role. So it's called someone because, you know, instead of calling out for Lauren, the people in their homes would call out someone turn on the lights or someone can, you, you know, lock the door. And even if you weren't sitting at a table, if you're in the gallery, you might hear the sound from it, you know, calling out for someone and you know, some little bit go to the table, hopefully and help them out. 
+
+James Parker (00:05:30) - What was it like fulfilling that role for you personally? And, you know, were you able to share it as sort of effectively that experience with the gallery goers? Because I imagine that there's something going on in terms of the temporality and plus know the proximity and intimacy that it's hard to reproduce. 
+
+Lauren Lee McCarthy (00:05:54) - Yeah, totally. I mean, I think when I did it, um, I was thinking a lot about control and the way that I had control over their home, but also in this commitment to be present with them and watching every moment that they were awake or around, they had a certain control over me, you know, like I'm, I would need to use the bathroom and I would like have to take my laptop with me in case they called out in the meantime, or they may be doing something very monotonous, you know, just watching TV or reading a book for them. That's a relaxing to be real for me. I'm just like, I can't do anything, but watch them do that. And just waiting for the moment, you know, an hour or two in when they're like, Oh Lauren, could you change the song? Or, you know, look up this word for me. 
+
+Lauren Lee McCarthy (00:06:40) - And I was really interested in the way that people felt this comfort in doing that. Whereas if the person was in your house, maybe you would feel more of a need to accommodate them. So I was like interested in that idea of like making myself into a system and trying to like have that carry through to the extent that that person felt that. So, yeah, and I, there was definitely like a relationship that developed really clearly between the two of us that felt very intimate and personal by the end then with someone you, I mean, you've named it exactly the challenges that it's not one person committed to watching this, like spending a week of their life, watching the person. It's a bunch of people coming through and maybe spending one minute or maybe, you know, some people would sit there for longer, like half an hour, but totally different. 
+
+Lauren Lee McCarthy (00:07:28) - Right. And yeah, they didn't have the same experience that I had necessarily because they, that like durational factor was not there. But I think that they got that feeling of, you know, I would notice people kind of sit down and play around with it. And then at some moment, see a person enter the frame and there'd be like this click of like, Oh, this is like a real, this is like real life that's happening. You know? And I'm actually the one, like the thing that I was just kind of banging on the interface, I was in an interface at someone's life or someone's home. And the other thing that I noticed was that even though there wasn't this continuity of one human behind the people in their homes really started to develop relationships with the system. And so that's where it got really interesting and murky for me because it was like, you know, I know people develop relationships with devices all the time, but it wasn't quite that because they're new they're humans there too. So it became this, they kind of had a relationship with the idea of a human behind this system without it being a specific individual. And so the, the devices in their home really took off, you know, someone in their home became this character for them, but I character was enacted or played by many different people over time 
+
+Sean Dockray (00:08:44) - Was thinking in the way that you were just describing it also. And also maybe there's something in the name, someone which is an acknowledgement that there might be no one, you know, like, like when you're calling out for someone, but deep down, you sort of admit that there might not be anyone on the other end. Um, which also seems to be a bit of a difference in that, like the responsibility that you felt personally to kind of being there throughout the entire duration is not necessarily felt yeah. In the someone in, you know, where it's a rotating shift of sort of more or less delegated or outsourced agents let's say. And so I guess what I'm wondering about is, is the role of like the failures, you know, like when someone's call would go unresponded to like, I guess like in, in, in those two different systems, there are the possibilities of like, um, their request being sort of like miss acknowledged or acknowledged wrongly maybe in one case, or like, you know, interpreted by a human, but in the other case, just maybe missed entirely and yeah. How the participants sort of, you know, calibrated their kind of like expectations, you know, like thinking about them as a performer as well. And, um, yeah. And it's those differences between those two systems? Yeah.
+
+Lauren Lee McCarthy (00:10:10) - There's a way in which we kind of comedy technologies, you know, we, we understand the ways to make them work, even when they're not behaving in the way they are necessarily designed. Um, I think we did that already with like apps and devices are like, Oh, you, and some of it is actually techniques for getting things to work again. And some of it is more like folklore, you know, like, Oh, if you do your thing thing with your phone here, and then that will work or whatever, um, like people kind of come up with their own theories and I've felt that kind of dynamic really strongly in both cases. So, you know, like there are moments where I realized when I was playing Alexa or when I was Laura and that I was like, wow, I'm just so much less efficient than Alexa, you know, they would say, can you turn on the hairdryer? 
+
+Lauren Lee McCarthy (00:10:58) - And then I'm like, Oh yeah. And I hit the button and I'm like, Oh wait, no, that's the, that's the faucet, wait, hold on. You know, like I'm kind of scrambling in a way that a machine wouldn't and to the extent that there were moments where I'm like, I remember one moment really clearly where someone asked like, Hey, did I take my medication earlier? I can't remember. And, uh, all I could really do is like, start to kind of, you know, I can't analyze all the footage, you know, on the spot, but I could jump around to different points and try and find it. And I was really aware of like this feeling of like, Oh, if I was an algorithm, I would just feel more confident in my answer right now. But all I can give you is just kind of like human guess or response. 
+
+Lauren Lee McCarthy (00:11:38) - And, but anyway, where I was going, I guess, was that the people on the other end were very, you know, sometimes you see people get frustrated at their technology or device, cause it's not doing the thing they want in this case. I think knowing there was a human error gave them this like patients just to be, so they were like happy to have that familiarity of like, Oh, I, that, that human humanness of not being able to get everything right. Or, or do the thing quickly, um, was present in the system. And then with the, someone, I think it was similar, but there was definitely more of a, like a time to adversarial relationship, you know, like points where they would be asking someone to do something like there's one moment where a woman was cooking and she's cooking in the dark. And she was like, it's someone, could you turn on the lights? 
+
+Lauren Lee McCarthy (00:12:30) - And there's obviously someone there, like, someone's manipulating the interface, but they're not turning on the lights for her. And you know, it just, it kind of added to the, it felt like an exploration of, um, what these personalities could be like, like there's a, it's funny, cause there's a part of this work where I had afterwards or in this process I felt like, Oh, like so many ideas for like what Alexa's personality could be like now or where this could go. Um, but I'm not, I'm very critical. Like I, you know, I don't want to go get a job working at Amazon, but it just like open, open my eyes to a lot of really interesting possibilities. I didn't really know what to do with after that. And so I think something that emerged through all these different performances and experiments were just seeing like range of ways people could relate to these things. I felt like much more open and creative and interesting than the way I think we often interact with these systems, which is very, um, structured. There's a certain amount of distance. He didn't with this kind of personal space that you're interacting with the device in. 
+
+James Parker (00:13:41) - Do you have any reflections on the agenda dynamics of your experience? I mean, I'm just ref just thinking back to a conversation, Yolanda stingers and Jenny Kennedy about that with the smart wife and they talk about, you know, the compliant that the model of femininity, that some of these devices embody, which is all about compliance and of course, domestic labor and that, that they also invite these companies to think about, you know, different versions of different kinds of personalities. And I just wondered if you felt strongly that that was a gender dynamic in, in your relationship. Um yeah. Or if there are any other reflections along those lines. 
+
+Lauren Lee McCarthy (00:14:30) - Yeah, definitely. I think I'm exploring that in these works, but it's not such a like pointed, you know, I'm not trying to put a point on that. I'm more kind of asking the question, um, you know, because it is a feminine voice in Lauren, which made more sense, I guess, because it's me, but I was also interested in that way that we like attribute, uh, a personality like Alexa or Siri to a woman, you know, and a lot of people have written about this, but I guess I was thinking about like, what is it about my femininity or my, the fact that I'm a woman that makes me able to embody this role in terms of our conception of what these systems should be. So it's was kind of playing with that and I was thinking, and so it's similar in someone to, again, it's a female voice and I think it maybe pushes it a little further because, uh,
+
+Lauren Lee McCarthy (00:15:22) - You know, it could be anyone of any gender sitting down at that table, but they are given this voice that sounds feminine. And so there's some question there about like, how do you fit yourself into that audio space and, and what does that feel like? And I think a lot of my work I'm interested in, there are a lot of pieces that I've done, where if I were a man doing them, they would be, they would have very different reading, um, or I might not be able to do them at all. Right. Because, um, like on one hand you are, you know, women are so much more often the, the target of, you know, the gays or, or just, uh, um, things like, you know, stocking or, um, being tracked in different ways or, um, not just thinking about gender, but thinking about race and religion identity in different ways, right. 
+
+Lauren Lee McCarthy (00:16:12) - There, there are some in the population where there's some people where, you know, for them surveillance feels almost novel to be kind of experiencing in this. Um, so explicitly, and for others, it's, it's much more of a daily reality, right? And so maybe they're not the ones that would opt to have this in their home or, you know, sign up for that, I guess I'm interested in. So there's the definitely like the dynamic of it, like who is privileged and who is not privileged within the system who is tracked and not, but I'm also interested in that, you know, by occupying this role, what does that allow me to see? What, what vantage point does that allow me to take? So it's not just about like who's being seen or watched, or who's more familiar with that tracking, but like, because you're not seen as a threat, um, you're seen as the target, like how do you, what does that offer you in terms of like, looking back? 
+
+Lauren Lee McCarthy (00:17:08) - But yeah, I did, I have done other pieces, so a more recent or another part taking the series, which is called waking agents was, um, the series of smart pillows. And so in that one, it was, there were performer dedicated performers for each pillow and people would like lie down with this thing and it would talk to them and it could play music. And with that piece, um, well, just to finish the description, the visitors were not told there's a human on the other end, they were just told this is embedded with intelligence and there was up to them to interpret that. And so most of them interpreted as machine intelligence and then there'd be some point in the conversation usually where they would have a moment of understanding or switching up their understanding where they would go from, Oh, I thought I was talking to a machine. 
+
+Lauren Lee McCarthy (00:17:51) - I realized I'm talking to a human. And so I was kind of excited by that, that moment of switching contexts, because it also means you go from feeling like you're alone to feeling like you are realizing that with another person the whole time. Um, even if you were never really alone, because, you know, who's knows what's on the other end of these technical systems. Um, but anyway, in that piece, um, the performance could actually choose and you could decide, do you want to a male voice or a female voice? I would love to, I think we're getting to the point now where you could also have like a gender neutral voice as an option in there too. So that was, that was interesting too. Cause they would, you know, people wouldn't necessarily pick the one that matched their gender, but I think for them, there was some questioning about like, why choose one or the other and what does that they would get to see in real-time like test it out. How does the interaction go differently depending on which of these voices they choose? 
+
+Joel Stern (00:18:43) - Um, the pillows are super interesting also because, you know, we'd been thinking about a lot about, um, wake words and, you know, and, and the way that sort of being awake or being asleep is, I mean, it's sort of an anthropomorphizing of the Machine in, in, in a, in a really sort of Lee literal sense that machines don't, you know, sleep or wake up. Um, but then, you know, obviously calling the work, Laura and draws, draws attention to the wake word and S the sort of call linked to the Machine, calling the Machine in, into sort of action. Um, but in effect you're always Listening and, and the wake word is, is not so much the call to wake up, but the call to sort of act on that Listening sort of in a, in a more, in a more transparent way. So I was just sort of thinking about if there were instances where the call to act sort of precedes the wake word, you know, where, where you felt I should act, I should do something, but I sort of, haven't been called I'm sort of not, uh, I haven't, you know, and then how that kind of dynamic plays out in terms of the ethics, because I think, again, this difference between human and machine in, in relation to sort of, um, w when it becomes.
+
+Joel Stern (00:20:05) - Sort of Lee legitimate to act when it, when you kind of authorized that. That's, that's an interesting question for us. 
+
+Lauren Lee McCarthy (00:20:11) - Yeah. That's so funny. I feel like I had this moment. I mean, I think I started by saying like, a lot of these projects are just kind of these attempts to try to hack my own, you know, social shortcomings. And so, you know, it's like, I've got it. This is my way in. Um, yet I get in and I'm still me. And, um, I think I had one moment in one of these performances and I was like, Oh my God, I just took all my anxieties and just like embedded them in the, you know, system infrastructure of someone's house. Now it gets out of control, um, all by like anxieties about like, should I act, or should I not? Or what should I say now? It's like, now it's distributed over your entire house. But the, I think that was always a question for me was like I do, yes. 
+
+Lauren Lee McCarthy (00:21:01) - Obviously, if they're they kind of use the wake word or they command me, then it's clear to respond, but it's, uh, it's a relationship that's unfolding when this performance happens. And it's funny because a lot of people go into it kind of thinking, Oh, it'd be like a show, I think, and then there's not much show, it's just like a situation that we're both in together and have to find our way through. And so how much I act beyond what is like commanded was really, um, it really different with each person. It was like trying to read them, but having very, having complete access to, you know, other camera feeds and information, but also being really aware of the times when you, that wasn't enough, but yet you're still trying to piece together a picture of them because that was the role that you chose to occupy. 
+
+Lauren Lee McCarthy (00:21:54) - And so I think for moments like that, it was like really ringing through my head. Like, what am I, you know, what we're doing when we're saying, Oh, well, we'll apply this algorithm. That's going to just figure out when things should happen. Right. It's, it's going to be an incomplete picture that we're then making these certainty or these judgements about, or the algorithm is making judgements about in terms of the, that kind of acting. Um, I think the other side of it was, there were times where I was asked to act and I, well, I guess one example was, um, someone had like a date over and, you know, she kind of said like, okay, this person's coming over. And like, can you kind of set the mood and lighting? And I did that. And, and they came and normally when someone else enters, I try to do something. 
+
+Lauren Lee McCarthy (00:22:41) - So they noticed my presence, you know, like, hello is your name or something. But like, they just kind of started getting right down to business very quickly. And so there wasn't like a moment. Um, and so I'm just like watching this thing and then I'm kind of like wondering like, is, you know, what does this person know? Or like, how am I complicit in this thing? Or is this, you know, am I just fulfilling my role here? And it's kind of her job to figure out what the boundaries are. So there were moments like that too, that were interesting and awkward. It was funny how that resolved, because at some point I was like, okay, well, I got to, like, I think I would like to leave them to this. And it was pretty late. So I was just like, you know, good night, everyone I'll be here tomorrow. And then at the moment he is kind of like what's happening. And, um, she was like, it's Lauren. Remember I told you about her. And he'd kind of, I think at that moment, like notice all these cameras everywhere and then they just kind of, then it was like, trug cool. Okay, great. Cool. 
+
+Joel Stern (00:23:44) - Most people probably don't care. I mean, that's, that's the sort of, um, punchline sometimes. Yeah. So were there, were there moments where, uh, I mean, that's a good example of a moment where you suddenly let's say couldn't pay the Machine. I mean, you had to sort of confront a human problem of making a decision about how completely you are in potentially intruding on this person. Who's come over without enough knowledge of what's going on. But I mean, um, without ever moments where the sort of you where the user or the participant sort of switched you off that Lee, like I said, um, I, I don't want, I don't want me, you know, because people, I mean, a lot of the people we've been speaking to in this project have been sort of thinking about this, the questions of how not to be heard and sort of coming up with strategies for, um, for doing that. So that kind of part of the relationship as well. 
+
+Lauren Lee McCarthy (00:24:44) - Yeah. And it's really important to me when I do. I'm always thinking about land consent so much in this work, and it's interesting cause it's like people can consent to things, but if you haven't experienced it before, you don't necessarily know how you might feel. Right. So how do you, how do you deal with consent in a situation like that? But one of the ways is just trying to be like, when I'm installing it.
+
+Lauren Lee McCarthy (00:25:09) - I work with it, with them to say, where, where do we put the cameras? Are there areas where you don't want them? And then also letting them know, like you can unplug the camera, you can cover it, you can turn it to the wall. You can tell me to stop watching. You can say Lauren, shut down and I'll turn off everything immediately. So they feel like, I mean, obviously there's a trust there, but I try to make that clear that's an option. And so then I noticed people interact with it differently. Like some people it's like cameras on don't care, you know? Um, other people are just like, you know, when they're going to sleep for the night or something, or when they're changing, they'll just like, turn the camera around or cover it or something like that, which I just thought it was like a really sweet gesture in some way of just like, yeah, I, I think each person kind of finds their own relationship with it. And I'm, I'm trying to like be there for that. 
+
+James Parker (00:26:02) - I noticed that you said camera a lot of times and that answer. And I'm just because, because the project is about Machine Listening. And one of the questions that we're interested in is like, to what extent are any of the questions that we're asking specific to Listening or audition? And, and the answer is often, well, you know, not so much. Um, but, but sometimes, yeah, I guess I'd just be interested to know if you have any reflections on the different modalities of being with the person and the different, like, so it seems that people were concerned about being seen possibly in their kind of nakedness or what have you. Um, but I always think about smart assistants that there's something about the vocality. So not so much the auditability, but the vocality that's producing something. That's that because of the longstanding connection between ideas of voice and intimacy and so on. So I just, I just wondered if you could reflect or sort of tease out any reflections you had on vision versus this. I mean, you know, even, I don't know, like memory or, you know, uh, different ways in which we, I think about the kind of sensory or affective dimensions of being with someone in these ways. 
+
+Lauren Lee McCarthy (00:27:17) - Totally. Um, it's interesting. I mean, I was saying camera kind of as a standard for like all around recording device, you know, because he cameras have microphones built in, but I think you're right. Some of those people that were turning them away, it was more about the image that they wanted to withhold versus the audio. But I think I've been like definitely gained a new appreciation for sound and Listening through these works. And I think in the waking agents piece with the pillows was one example that kind of came out of that because it's all about audio because I mean, first of all, it's just like bandwidth wise, it's easier to deal with, but I found myself like in these performances because I'm, I'm building the software and the technical infrastructure as well as performing them. Uh, there's always like some learning and some development that's happening along the way. 
+
+Lauren Lee McCarthy (00:28:12) - And I found myself like picking up these different modalities to try to get clues about what was happening. You. So sometimes the camera does not give me any information. And so I'm relying completely on the audio to try and understand what's happening in the home. Yeah. Either the it's not useful or it's distracting because it's of the quality or, or whatever's happening in it or the lag, but also because it just, so that that's the incoming, right. If I'm to observe, but then as you mentioned, the outgoing and I think about these systems so much more in terms of that kind of audio experience of just like having your house and the fact that there isn't any screen normally. I mean, I know there's some, you can get like Alexa touch or whatever, but like the idea that it's just audio and it's disembodied makes you feel like it's everywhere, right. 
+
+Lauren Lee McCarthy (00:29:02) - Instead of just one limited to one device, I definitely played with that in the performances a lot. And I would have a few different speakers that I could like Bluetooth connect to and switch to different spaces of the house too. So I was often following them with my voice. Um, and yeah, and like, as I mentioned, this smart goal, a piece that was all about audio and I I've really found that. I dunno, I feel like there's a way you can like understand the space you're inhabiting so much better just by listening to the audio then by looking at the image sometimes. Um, and that was, that was something unexpected for me. Cause I think it hadn't been such a big part of my practice previously. And I think for them to on the other end, like, um, the thing that really brought that sense of intimacy was the voice. It wasn't, I could control all the things and the lights and the appliances, but it was the conversation or the voice and it, and I use conversation really broadly. You know, some people we would have long discussions and other people, it would be very limited, but I think it was that continuity of audio that built the character for them.
+
+Joel Stern (00:30:10) - I mean, there's also such a rich, um, sort of cultural history of the disembodied voice, whether it's in radio or, or in cinema or, or you think of CA voice characters, like how from, from 2001, and, um, the kind of omnipresence of, of the disembodied voice, like by specifically, by not having a body, it can be everywhere at once and how have a certain, you know, power that comes from being located sort of nowhere and everywhere. And also, I guess it's, it's, you know, the way in which that voice produces a sort of a subjectivity, even though the voice might be Machine  ...  sort of, it feels like it's listening to you and it feels like it's speaking to you. Um, even though sort of listening and speaking might be human rather than machining sort of qual qualities in a literal sense. It, you know, the Listening is audio sort of processing and the speaking is this sort of form of synthesis, but, uh, we, we experience it as if we're interacting with a human and there's a certain power. 
+
+Joel Stern (00:31:25) - I mean, one of the things that's maybe interesting to submit a sort of think about in this cultural moment would be, you know, the way in which the voice of sort of Siri or Alexa, or in this case lower, and it is sort of, you know, friendly and, and helpful and understanding, whereas the voice of say, you know, how, or, or previous DC embodied voices have a certain malevolence as they become conscious of their own sort of power and agency. So I'm just wondering if that, you know, and I guess that's one of the things that's in this project too, is, um, this with sort of thinking about the spectrum sort of from utopian to the dystopian, her horizons of Machine Listening. So, I mean, I feel like your project is sort of broadly optimistic in that it's sort of strongly humanistic, but I just wonder, um, whether there have been moments where you have really had a sense of, you know, not so much the, the, the, the dangers of these forms of mediation, but the more dystopian sort of horizons. 
+
+Lauren Lee McCarthy (00:32:35) - Yeah. I think, I mean, these DCIS are never, they're meant to be spaces for people to try to sort out some of their feelings or responses to the technology. Um, I have my own perspective, which is normally quite critical. Um, but I'm trying not to impose that, um, and to let people find their own relationship. And also that, I guess for me, there's always some part that's like a little bit about hope or at least some pleasure, um, because it, I think if it's, if there's not that that people just shut down and I don't know how actionable that is, um, if we want to actually imagine, uh, a different world, but that being said there, like in every performance, there are moments I think, well, I'm aiming where you feel some real sense of connection and maybe in a way that you hadn't anticipated, you know, and that's kind of the hope of like, Oh, there are this, these that's something different, but then there are these moments that feel, I think, incredibly dystopic or, um, I've had some people interact with different pieces and just be like, wow, that was like a depth of darkness. 
+
+Lauren Lee McCarthy (00:33:50) - I didn't know. I could go to an art piece. Um, and it's interesting cause the moments are not, they're not necessarily, you know, like there's the time where I like, uh, I messed up the kind of power rating as I was like translating this piece to Europe and then the whole house went dark and I had this moment where I thought like, maybe I just like burnt down their house and you know, like who, let me do this. And who's insuring, um, right. But those aren't necessarily the darker, the more to still pick that they're often in the kind of quieter times for a year, just like, Oh, I feel close to you. And I feel this like huge distance right now because of this remoteness or like, how did we end up in this? I know it's an art piece, but like, how did we end up here? Yeah. And I think those are the same things we feel when we use technology to, um, but maybe you don't have so much. I think the thing about feeling those while you're scrolling through your phone or whatever, is that the whole system is set up to keep you just moving past it. Whereas I guess what I'm trying to do here is just like, sit with that for a little, for a minute. I mean, you're also not a profit extraction Machine. Unfortunately. 
+
+Lauren Lee McCarthy (00:35:06) - Well, artists, artists are, you know, which has extract little more profit.
+
+Lauren Lee McCarthy (00:35:15) - Yeah, definitely different goals. Right. And that, I think that's a big part of why the relationship feels different. I mean, yes, there's a human on the other side, but there's a certain trust that you just, I think is impossible to have with some of these other technologies. 
+
+Sean Dockray (00:35:28) - That's a quick thing that I was thinking about is just, um, on the one hand, like I'm there, whether it's you or gallery goers kind of like sitting in for the role of, uh, you know, corporate personality, you know, substituting for the algorithm, but then at the same time, there's the like, you know, massive kind of actual human labor sort of underlying a lot of the algorithm, not just the algorithms, but sometimes they kind of like actually are performing that, that role. And so obviously the pieces seem to relate to that in many ways. I just wondering if he'd say a few words just, um, on that, cause I'm sure it's something you've thought about in the course of making the works. 
+
+Lauren Lee McCarthy (00:36:13) - Yeah, definitely. Um, and it was, uh, I'm glad you brought that up again. Cause I was actually starting to think about that when you were talking about like the image of versus the sound or just like what's seen and what's not seen. Yeah. I mean, everything from now, you just dropped the link to the, um, you know, what's called artificial urge, artificial intelligence, right? Like a mechanical Turk, that's kind of their catch phrase. But you know, I think after I started doing this project, like there was a news headline. It said like there actually are humans listening to Amazon Alexa. And when further to talk about what that experience was like, which is, they're mostly doing quality control, but they're, you know, often overhearing things that could be extremely disturbing. There's not a lot of guidance in terms of like what to do in here. 
+
+Lauren Lee McCarthy (00:36:58) - Those in this article mentioned like, what should you do when you hear that? Well, go to this like chat room where you can kind of like commiserate with other people that are also feeling doing this work and feeling traumatized. That seemed to be like the response. So yeah, that's, it's, it's horrifying, you know, it's every, it's on every system that we're interacting with. You know, there's like a great documentary about, um, the workers behind Facebook, you know, that are filtering the content. And I don't think there are a lot of good answers in terms of like, what do you, what is the way to address that? And it's, it's so such a big problem and it's so invisible. And so I dunno, hopefully there's something in this piece where people think about that aspect of it a little bit too. And also like, what does it say? 
+
+Lauren Lee McCarthy (00:37:49) - You know, there's a specific audience that is experiencing these artworks. Um, as much as I'm interested in like reaching a lot of different people, it's like they are the people that would elect to sit behind this desk and do this as like a novel experience. Right. Because it's far from the, the work that they do on a daily basis, you know, there's a huge sector of the like care work is the fastest growing, one of the fastest growing job sectors, at least in this country. And how do we, how do we even think about that? You know, or address like what a big need that is as we're also like on this fast track to automating as much as possible. I think we'll start to realize there's something you can't automate and are those jobs seen as, you know, things that we value or is that, um, human labor things that they try to be invisible completely. 
+
+James Parker (00:38:41) - I wonder if that's a good note to end on, um, could open up a whole new thing, uh, and I'm tempted, but, um, um, but it was so interesting. Um, thanks so much, Lauren. Thank you.

content/transcript/parker.md

@@ -0,0 +1,126 @@
+---
+title: "Angie Abdilla"
+status: "Auto-transcribed by reduct.video"
+---
+
+Jasmine Guffond (00:00:21) -  ... , this is computational Listening with Jasmine Guffond. This first edition includes an interview with academic writer and curator James Parker, about his research into Machine Listening, followed by a selection of electronic music and track listing is available and joy. Hi James, thanks for your time this evening. 
+
+James Parker (00:00:58) - It's a pleasure. 
+
+Jasmine Guffond (00:01:00) - So you're a senior lecturer at Melbourne law school, which is part of Melbourne university. And you're also an associate curator, uh, with liquid architecture, which is an Australian organization that curates artists working primarily with sound. 
+
+James Parker (00:01:17) - Yeah, so I mean, you know, there's not that many people in law working with sound and Listening. Um, so, you know, I I've carved out a little niche for myself. I wrote this book about the trial of Simon Bikindi, um, who was a Rwandan singer musician popular figure who was accused of inciting genocide with his songs. So what I was trying to do in that book really was to just to think about, you know, what would that really mean? What, what would you have to do in order to think seriously about the possibility that someone had incited genocide with his songs? And what I say basically in the book is that the international criminal tribunal for Rwanda doesn't really take seriously the kinds of questions about the nature of sound, about the nature of music or Rwandan music or random music on the radio in a specific Listening context that it would need to in order to, you know, to seriously grapple with that question. So, you know, subsequently moved on to think about law in relation to Sonic weapons and the weaponization of sound. And then subsequently in my work with liquid architecture, I've been working on eavesdropping. So I kind of thinking about the law and politics of Listening and being listened to, and, you know, particularly in the kind of surveillance contexts. And that's what led me to my most recent work on Machine Listening, which is sort of only just getting underway now.  
+
+Jasmine Guffond (00:02:37) - Okay. Could you explain what you mean when you use the term Machine Listening and perhaps give one or two concrete examples? So for people who haven't heard of the term before 
+
+James Parker (00:02:48) - Yeah, sure thing. I mean, I think in some ways everybody's kind of familiar with, you know, the idea of Machine Listening, anybody who's engaged with a voice assistant or a smart, you know, or even a call center where, you know, it asks you to say your name or respond yes or no, or say a number or something. Right. So, you know, Machine Listening on one level is just engagement with a Machine that seems to understand you, right. And they're getting increasingly good at understanding you. So, you know, that's, that's one sort of simple way of thinking about Machine Listening. It's, it's engaging with machines that seem to listen, but actually it's a term that's being used in the scientific literature in, uh, as well. So there's a sort of an increasingly large number of researchers working on audio and AI basically. And they sometimes refer to what they're doing is Machine Listening. They also use a whole number of, uh, synonyms. So people might be familiar with automatic speech recognition, right. Which is the kind of very sort of speech oriented version of this, which is one of the oldest disciplines of Machine Listening actually. Um, but there's also sub fields like audio event detection, which is about training, um, algorithms, neural networks, machine learning systems, to be able to understand Sonic events like the shattering of glass or a crying baby or a gunshot, or, um, what have you and audio go in 
+
+Jasmine Guffond (00:04:24) - And would that be used to try and say locate you? Like what room that you're in or if you're in a shopping mall or at home, or, 
+
+James Parker (00:04:31) - Yeah. So, um, from my understanding, uh, what you're describing is more like audio scene analysis. I mean, I don't want to get too technical because I think in some ways they overlap and, you know, in order to understand a Sonic space like a mall or something, you need to be able to detect speech for example. And so that, you know, th th it's not like there's no speech analysis going on in audio scene analysis and so on and so on. But yeah, audio event detection is about identifying very specific sounds. So one company, for example, audio analytic, which is a UK based company,
+
+James Parker (00:05:04) - Sells various security systems that can determine, for example, you know, the, the sound of a, of a window shattering and that, and the idea is that then they can trigger an alarm or, um, trigger some kind of automated system that would alert police or, um, alert the user via this, you know, their smartphone or what have you, you know, that they have a product that, um, would be in cars, right? So that it would be able to hear the sound of a bell or an engine or a driver yawning. So audio event detection is about identifying very specific sounds. Whereas audio scene analysis is about recognizing ambient environment and then sort of determining, automating what to do with some kind of an ambient environment. So audio scene analysis, the kind of thing that you have embedded in your smart headphones. If you've got smart headphones, maybe you're sensible not to, uh, 
+
+James Parker (00:05:53) - But they're, they're not devices there, there are, um, Headphones now. Um, you know, so headphones will once for Listening and now headphones listen in order to supposedly improve the Listening experience for you. So they can tell whether you're listening in a noisy environment or a quiet environment and adjust the noise canceling accordingly, or they can tell whether somebody is trying to speak to you, or this is the claim anyway, and then they can turn on ambient and microphones embedded in the, in the headphones that then can sort of transmit the speech of the person speaking to you via the microphone. Now, like how useful that is, how much it improves the Listening experience, whether it's just, um, another way of making a product smart in order to extract more data and more information. That's a slightly different question, but so to return to your original question, you know, what's Machine, this needs, it's a field of science and technology that people are increasingly familiar with via things like smart speakers, voice assistance, but it's much, much More than that. Smart headphones, gunshot detection, voice prints, emotion detection. I don't know if you read about this new device launched by Amazon, Amazon halo, that's kind of like a, like a Fitbit, but that listens to you as well, and supposedly tracks your emotions throughout a day in order to give you feedback on whether you were particularly happy or sad or depressed, or what have you, and then  
+
+Jasmine Guffond (00:07:23) - Via the vocal quality from listening to your voice. Yes.  
+
+James Parker (00:07:28) - Supposedly, and there are, I mean, there's all sorts of other, uh, similar applications, aggression detection for security systems. The idea is that they'd be able to identify in advance, you know, an argument as it's brewing from the sounds of voices in relation to a sort of normal, um, ambient environment, also age detection, gender detection, supposedly ethnicity detection, depression detection, outsiders diagnosis, one company, one American company, um, does vocal risk assessment. You know, what if they think risk is, is a fascinating question, but, you know, they, they claim be able to, and by means of a two to 10 minute long phone call, determine whether or not you're, um, a risk to the company that's buying their product in some way in military context. So as a screening mechanism for taking on contractors or, you know, employing, you know, soldiers or, um, security agents, or what have you, but also, you know, they're selling their products to banking, to detect supposedly frauds all from the sound of people's voices. 
+
+James Parker (00:08:32) - You know, so there's many, many, many uses or putative uses of Machine Listening techniques and they're growing by the day. And I don't think that the friendly face of, um, not that they have faces of things like Siri and Alexa, they take up a lot of the attention where I think actually this kind of field of Machine Listening has growing and making itself ubiquitous more than people tend to realize. And so for that reason, I think we should think of Machine Listening, not just as a field of science and technology, but also of as a kind of a field of power, a field in which politics is happening. You know, that's connected to other systems, exploitative and oppressive systems. You know, it's obviously collected in some way to capitalism, to data colonialism, you know, even white supremacy and patriarchy and things, right? So this is a, an emerging system of power, not unlike computer vision, not unlike facial recognition. Um, when people think about the politics of, you know, search and other algorithms, they might begin. I hope to think about the politics of something like Machine Listening too, because I think it's a bigger deal than people are giving it credit for at the moment. And it's going to keep it. 
+
+Jasmine Guffond (00:09:47) - We are certainly painting a dystopian image. I think we're at each and every one of our devices are listening to us and then using that data to predict how we might act in the future and in a way sort of assuming that.
+
+Jasmine Guffond (00:10:03) - We could be acting negatively like aggressively. I mean, I imagine there must be so much room for error. Like you could just be annoyed because you stubbed your toe walking down the street and is that going to then be misconstrued? 
+
+James Parker (00:10:16) - Well, that's one of the interesting things about this. So to the extent that any kind of AI system is political both to the extent that it works and there's a kind of Omnisphere and surveillance system, you know, that knows too much and knows too well. And just as problematically to the extent that it doesn't work, to the extent that it embeds and, uh, normalizes and makes seem neutral or objective existing biases, like racial bias, gender biases, and many, many others, but also in trenches and produces kind of new and more arbitrary biases and errors of its own. Right? So the problem comes both ways. It's a challenge politically, to the extent that these systems work, we might want to think about whether we want the systems to work and the way that they claim to. And that is a challenge to the extent that they don't work. And sometimes what we're being sold is simply snake oil. You know, that the promise of AI as a sort of a marketing term, outstrips the actuality. And so I think that there's many companies that are claiming to be able to do things that they simply can't do and with very little kind of scrutiny and regulation. 
+
+Jasmine Guffond (00:11:37) - Yeah. I mean, it seems like one of the issues with AI is its inability to be able to tell context. So am I hear the sound of breaking glass, but that could just be you knocked over a glass at home. It's not someone breaking into your house for example. 
+
+James Parker (00:11:53) - Yeah. But the problem with that argument, that what some of the problem with your argument is that you can just imagine like the cogs turning in the AI technicians mind, right. Because the immediate answer is, well, what we need is we need more context 
+
+Jasmine Guffond (00:12:06) -  ...  or more data to define context. 
+
+James Parker (00:12:10) - Right? Exactly. And it's the same when people point out biases. Right? So for example, there's a great scholar in the U S called Halcyon Lawrence. Who's written about the racial biases of Siri and other voice assistants and the way in which they are unable to, or sort of refuse to hear accented speech and the way that, that taps into a long history of oppression and imperialism via and in relation to language. And there are all sorts of questions about what, what you do about that problem. So there are obviously not financial incentives for Amazon or Google to go after minority markets, right? Minority speech markets. You know, you might say like a heavily accented speech from whatever communities, but you know, if they were doing AI for good, maybe, maybe they would. Um, and so, you know, you have this logic where you go, well, let's make it so that these assistants can understand all forms of speech as being inclusive. 
+
+James Parker (00:13:06) - So, you know, a politics of inclusion that's on neverless that wants to understand everything better and perfect itself constantly, you know, it tends towards there being no limit. So the kinds of data or the people and the places and the context that warrant data extraction. So there's a bit of a sort of a trap in, you know, pointing sometimes to the errors and the problems with the systems because the immediate response is, well, we can correct that if only you you'll allow us to have more data, but what's never questioned is the frame where whereby we need the system in the first place. 
+
+Jasmine Guffond (00:13:46) - Well, speaking of frames and no limits, would you be able to give an example of Machine Listening, being used in health applications? So particularly in the current context of the COVID-19 pandemic, 
+
+James Parker (00:14:00) - I can have a go, but I guess I want to premise what I'm going to say by pointing out that I don't really know to what extent these things are actually happening already. So one of the things with the pandemic context is that, you know, things are happening extremely fast. I mean, it's clear that big tech is using the pandemic as a way to sort of expand its tendrils and tentacles into more and more spaces. So one of the moves that you're seeing big tech do is move into health. You know, all data becomes health data after COVID, you know, um, and the same way that big texts moving into education, uh, in homeschooling and home universities, there's an opportunity, a market opportunity here. So, so people are trying to move in to the health space. That's that's really, really clear. And one of the ways in which they're trying to do it in the Machine Listening context has been COVID voice diagnostics. So this is the idea that we could tell if we just had enough data and trained our algorithms.
+
+James Parker (00:15:05) - Correctly that you have COVID based simply on a voice sample. So it might be your cough, or it might be the, you know, a sample of speech. And, you know, this responds to the intuition that we have that, you know, you can sort of tell when somebody's got a cold, right. You know, they don't seem to be speaking. Normally they sound stuffed up. The idea is that, well, maybe even though I couldn't tell you whether you specifically have COVID, you know, just sound a bit funky wouldn't know, but maybe the machine learning system can tell the truth of your voice, you know, beyond the limits of human Listening. And so there's all sorts of companies and universities. Um, and then, and entities that have tried to move into this space. And the reason I was a bit cagey at the beginning is that that's not at all clear to me, which if any of these things work, to what extent they work, whether they're actually being deployed already. But I think we're going to see it happen pretty soon. So to give some examples, there's an organization called voca.ai, which is ordinarily a, um, they provide this an Israeli company that provides voice assistance in sort of call center context. So they're sort of AI driven voice assistance. They partnered up with some researchers at Carnegie Mellon, I think by March to gather data in relation to, to train a system in order to detect 
+
+Jasmine Guffond (00:16:31) - COVID. So they're jumping right in there 
+
+James Parker (00:16:34) - And right in, they were sort of ready to go. And so they were getting hundreds and hundreds and thousands of samples really quickly that they, they had all this language like volunteer your voice, you know, to help, uh, fight against COVID-19 
+
+Jasmine Guffond (00:16:48) - Actually, I'll probably play that Sonify ad so that listeners can hear an example if, um, so the company's asking for people to voluntary, um, give over voice recordings for their data set to train there. Yeah. 
+
+James Parker (00:17:03) - And then this idea of like voice donation, I think is really interesting. And, you know, with voca.ai, I don't want to sort of say anything that will get me into trouble, but like, I think I can stay at least that it wasn't, it's not clear that they won't use your data to train their call center agents, you know, uh, uh, beyond the, um, you know, w w whether or not anything comes out of the COVID voice diagnostic attempt. So that's one, that's a private organization. Cambridge university has got a team working on something like MIT has got a team that, that one's a bit different. They, they were trying to train their algorithm on newsreaders that they gathered from YouTube. So they had recordings of newsreaders speaking once they'd had a COVID diagnosis, and then they went back and trolled through their history and found recordings of them prior to getting COVID. 
+
+James Parker (00:17:55) - And then they sort of, you know, did it compare and contrast, and, you know, they say that they produced a study and says, Oh, it's very promising and so on and so on. And then there's companies Sonify, which are claiming that they've already, they're already able to do this. And the Sonify ad, you know, um, um, says that they're, they're seeking FDA approval right now. Yeah. Uh, no, I find that really interesting. I want to know, I just want to know what's going on, uh, what paperwork and what conversations are being had. It's not that I don't think it's conceivable that COVID voice diagnostics could work, but, um, at least a little bit skeptical. And I'm also really concerned about, you know, what systems such a thing would be embedded into and who's scrutinizing it and what measures there are for false positives. And then my mind, you know, you said before that, you know, it's just opened my mind. 
+
+James Parker (00:18:54) - I think I've got a sort of a dystopian tendency. And I just imagine a world in which your health is always being monitored by means of your voice. And it's not a fantasy. Google has it in patents, Google already has, or maybe it's Amazon actually has a patent, you know, it's Amazon because they have a pattern that so a few years old now that recognizes you coughing when you're speaking to Alexa, and then it begins to offer you, would you like me to, you know, buy some, some cough medicine and get it shipped to your house? And then you can imagine a world, right? Because you know, one of the promises of voice diagnostics like this is that they can tell you have the thing before you, before the symptom is kind of fully realized. So before you're even coughing, you know, it might be able to hear that kind of the pre cough. And, you know, you could imagine a world in which, you know, the cough medicine arrives before the cough, uh, you know, it's a kind of, uh, you know, that's the end point of logic. So, you know, what would it mean to live in a world in which it's not just that the content of one's speech is being monitored, but the, the manner of one's speech, the manner of one's health, you know, it's, it's like, you know, the embedding of.
+
+James Parker (00:20:03) - A stethoscope into every surface and environment. I don't know if you saw recently that Amazon's just launched this thing called Amazon residential, or maybe it's Alexa or echo residential or something whereby they're sort of working with landlords to embed, um, Amazon echo and Alexa devices sort of throughout entire apartment complexes 
+
+Jasmine Guffond (00:20:26) - And next to the police. Right. Is that, did the police get it? 
+
+James Parker (00:20:30) - Sure. To be honest. Yeah. You know, um, uh, that, that one, I just read like a couple of days ago, but, you know, sort of on one level the details, I don't want to say that the details don't matter of course the details matter a lot, but Google nest, you know?  
+
+Jasmine Guffond (00:20:47) - Yeah. Google does. And Amazon has one that does as well, like, um, is it called neighbor? I forget now, but they definitely, 
+
+James Parker (00:20:55) - You know, the, the, that that's the, that's a horizon at the very least the sort of the system of power and control whereby like walls monitor your health continuously. And they have a continually expanding ability and to not have really control over because you never do, uh, over what, you know, what kinds of things these devices or these ambient systems know about you or capable of interpreting, you know, I think that's, that concerns me. 
+
+Jasmine Guffond (00:21:33) - Yeah, totally. Um, I'm just gonna quickly jump back to FDA because the first time I saw the Sonify ad, I actually had to ask someone what's the FDA. So it's the federal drug authority. Is that right? Yes. Okay. Yeah. 
+
+James Parker (00:21:47) - And I, I have no idea what ability and organization like that has to think about political questions in relation to sound and AI systems. So who knows what's going on there 
+
+Sonify Ad (00:22:00) - COVID-19 is potentially the single greatest problem facing humanity today. The inability to rapidly, safely and accurately test for COVID-19 is one of the most challenging factors that has led to its infection rate. Sonify has developed a voice detection algorithm using machine learning and AI to identify very specific health characteristics in the human body. Our team of scientists and technologists have found a way to tell if a person has COVID-19 by simply analyzing their voice on a mobile device. Since we have identified the biomarkers of COVID-19 in the voice, we are now working with the FDA to seek clearance for our technology and need your help. We need to provide the Sonify machine learning algorithm with more validated voices of people positive with COVID-19. If you or someone, you know, has recently tested positive for COVID-19 within the last two weeks, sharing 30 seconds of your voice at  ...  dot com slash R. I can Help us get this app to the world by sharing your voice, you can save lives. 
+
+Jasmine Guffond (00:23:13) - So one of the concerns that emerges from your research that I'm particularly interested in is that Machine Listening bears I'm quoting from you now little relationship to the biological processes of human audition. So for us to really comprehend how machines listen, we need to move away from thinking about Machine sensing in terms of anthropocentric modes of perception. Like they don't listen in the way we as humans listen. And so you've come up with a couple of terms to explain Machine Listening and how it's different from human Listening, such as Listening effects and operational Listening. If you could explain what you mean by Listening effects and operational Listening, and also, do you think it's misleading to continue to use the term Listening in relation to machines? 
+
+James Parker (00:24:02) - Hmm. Oh, there's so much in that question. It's a great question. So where should I begin? Just to unpack the question a little bit, you said one of my concerns is that Machine Listening bears literal relationship to, you know, the biological process of humans. I just want to clarify 
+
+Jasmine Guffond (00:24:18) - That emerged. 
+
+James Parker (00:24:19) - Yeah. I just want to clarify that I don't, it's not the fact I don't care. It's not like they should, it would be better. I'm not sort of, you know, a humanist in that sense. Um, but I think we need to understand that that's what's happening. Um, you know, actually the early kind of experiments with automatic speech recognition, we're really trying to ape, uh, and replicate human methods of Listening. And there's a great scholar called Xiaochang Li, in the U S at Stanford currently, who's done. Some work on the history of automatic speech recognition. And one of the things she talks about is how it was the kind of the abandoning of trying to model human Listening in favor of statistical modeling that kind of provides the break through IBM thinking the seventies, although it could be wrong about that. I can't quite remember. That's the breakthrough moment when we abandoned trying to listen like a human word, you know, so, okay. I could be a bit flippant, you know, what's the problem with the frame, the framing, but framing this, all of this in terms of Machine Listening well, um, it's not really machines doing it on one level and it's not really Listening either. So why talk about it in terms of Listening as a matter of politics or sort of advocacy? I think there's something helpful about thinking in terms of Machine Listening. 
+
+James Parker (00:25:36) - So I think people understand what you're talking about and if you swap out computer audition, you know, which is another term that's used in the scientific field. I don't know if people, people are probably going to translate it into something like Machine Listening in their head. Anyway, I like the fact that Machine Listening kind of sounds a bit like Machine learning and machine learning has already got kind of some political traction so on. And so it's an analog to machine learning, but at the same time, as you identify it as a political problem, I think we immediately need to move on to say, okay, it's not working in exactly the same way as, or anything like the same way as human Listening. And that has important political what technical and political consequences we need to follow those and see where they lead. 
+
+Jasmine Guffond (00:26:24) - Okay. I was just going to say, the reason I ask is because I agree with you. I think it's really important to understand the nature of Machine Listening and that it's not, um, it's know essentially data extraction because there's one great example when, um, for me, when Zuckerberg spoke before the us Senate about Cambridge Analytica. Yeah. And, um, he was asked by, um, I think it was Senator Gary Peters. Do you use microphone to obtain personal information about people and zackerburg could very easily just say, no, we don't. And he also said, that's this like conspiracy theory that gets passed around that we listened to what's going on with your microphones. And so he was kind of able to sidestep the fact that they do gather so much information about us. Um, and they don't need a microphone necessarily to do that though. Of course they do also have patients to use microphones. And that's why I think it's really important that policy makers and politicians at least understand the nature of Machine Listening, because otherwise they're not able to ask the right question. 
+
+James Parker (00:27:34) - Yeah, no, I, I agree. Um, I don't know if I understand the nature of Machine Listening. I mean, you know, it's like with computer science, it's sort of a grab bag of techniques, you know, in ones and one level, I'm trying to name something that say political or so-so socio-technical object, uh, sorry, that's a bit jargony, but a social and a technical object, as well as, you know, rather than just simply a technical one. I want to say that Machine Listening is as a system of power, rather than just a technique of Listening. It's a whole range of techniques that intervene in the world by methods that are both analogous to Listening and the sense that they use auditory data and they sort of comprehend or analyze it in some way and are experienced as this to me. Um, so that's one of the things that I was trying to get at when I, when I talk about Listening effects, right? 
+
+James Parker (00:28:31) - So that it is really important to focus on what precisely the Machine is doing. But it's also really important to understand that there's, it matters that we experience ourselves as being listened to it matters that one moves through the world and feels that the, you know, the, the walls have ears or doesn't feel enough that they have is, you know, and one way of responding to the question just to say, well, we could, we could also what it means to, to be in the world and feel that, you know, you're being listened to. That's important too. When I talk, I talk about operational Listening. I mean, and I should say that I borrowed that in the first instance from a guy called Mark  ... , who's an amazing, um, American scholar he's based in Melbourne right now. Um, he's just written a book called automated media and he's borrowing it in turn or expanding it from Harun Farocki and Trevor Paglen, his ideas of the operational image or operational vision. What broken Paglen and getting at is the idea that images are increasingly being created for Machine eyes or on the visible at all to Machine eyes, which are not eyes. So there's a whole world of images being created, which totally bypass.
+
+James Parker (00:29:54) - Human perception that are created by machines for machines. And so Mark Andrejevic, which says, well, that idea of operationalization where the, where the Machine, where the image production or the sound production is just that to do something within a system, it performs an operation. It does something right. It's not for aesthetic perception or consideration. There's no understanding. It's just an image is produced. It performs, uh, an effect in a Machine X system that apprehends it on some level and, and, and therefore produces some kind of result. Well, we can draw that out in relation to Listening too, and say, that sounds, uh, increasingly operationalized, right? That they're not there for understanding Listening is not like a relevant consideration. So if we say operational Listening, we're meaning a form of Listening that is trying to do something rather than understand something. It, that when Amazon echo listens to you, it's not Listening.  
+
+James Parker (00:30:55) - It's just trying to produce some kind of result, which is often going to be to try and sell you something, you know, so it's a kind of a Listening without understanding or a Listening, a purely correlative Listening, as opposed to kind of comprehending Listening. It's a Listening that is kind of is just an operation within a Machine X system. And so that's a way of responding to the same question that says, okay, well actually we do need to think about the sense in which this is not really Listening and the way that we understand it as kind of comprehension or knowledge it's, it's bypassing that it's, it's something different. It's there are sounds being apprehended in a way that's quite particular and exclusive to machines and increasingly sounds being made by machines that are only for machines as well. I don't know if you've heard about adversarial, audio and adversarial. 
+
+James Parker (00:31:50) - Audio is basically audio that is intended is produced by a typically some kind of algorithm. Um, I think exclusively by an algorithm that's meant to be, apprehensible only to some other Machine Listening device. So basically the reason it's adversarial is that you can play audio. That just kind of sounds like noise to a human, or maybe it sounds like part of music, you know, a piece of music, or you can sort of overlay it or embed it in, you know, your next album, Jasmine, uh, and, uh, but, um, you can trick the Machine Listening system in, in, you know, a smart speaker or some other kind of system to understand it as either, you know, a trigger, you know, so there are examples, if you Google adversarial audio, you'll see these, you know, you can click play and it will, it'll go. Th that shows you what the Machine is understanding as being, as having been said. 
+
+James Parker (00:32:52) - So for example, you know, buy me five packets of whatever and deliver it to my house immediately, you know, um, or email such and such or, um, whatever it might be. So these are in other words, sounds that are the content of which is not available to human ears, but which are audible to machines. And so that can be used to kind of, to intervene in the sort of the soundscape in a way that sort of bypasses human comprehension. So it sounds made by machines for machines, right? So, um, and you know, obviously there are espionage and kind of hacking kind of CA capabilities there. 
+
+Jasmine Guffond (00:33:34) - It makes me think of the, you beacons, the ultrasonic beacons, and that's, um, where your TV or radio, or say YouTube content will emit an ultrasonic frequency above a human hearing range. And if you have an app on your device that can receive it, it'll receive it and then send that data back to the advertising company. Usually what you're watching and what time, and it's also, 
+
+James Parker (00:33:58) - This is not ultrasonic, but you're quite right. Like it's doing that, it's doing something similar, isn't it? 
+
+Jasmine Guffond (00:34:06) - So it's a means of communication. Yeah. And it's also used apparently to follow you around supermarkets and sports stadiums. Right. But, um, I think you have to go, so I was gonna just ask you jump to the last question, because over the course of these three radio shows, I'm asking different kinds of practitioners about strategies for addressing, um, ubiquitous and pervasive modes of contemporary surveillance. So how would you propose to challenge invasive Machine Listening or what would be an approach I've spent like the last 
+
+James Parker (00:34:41) - Few days, uh, trying to think about that exact question in relation to a project that I'm working on at the moment with liquid architecture, that's going to be launched at unsound in October. There's a part of it called lessons in how not to be hurt. And so I was trying to write that over the last couple of days, but I'm going to bracket that question because what I ended up saying in that section is like the sort of groundwork to be done. First of all, we need to, we need to identify Machine Listening as an object of political contestation and artistic and aesthetic and activists.
+
+James Parker (00:35:13) - Inquiry and, you know, and to some extent it is, but, you know, comparing it to something like facial recognition or computer vision, I mean, we're way behind. So one of the, one of the purposes of the unsound project is to sort of begin to grow a community of people, activists, academics, artists who are, um, produce Machine Listening as an object of concern, right. Um, and to produce a kind of a group of a network around it. And so we're calling the project with unsound Machine Listening, our curriculum because we're involved in, uh, you know, we're, it's early days for the research. So for us, it's kind of like the beginnings, you know, w w we're studying it now, you know, live in real time. We don't have the answers where we want to work with people like you and others to begin to produce the answer to those questions. 
+
+James Parker (00:36:03) - So it's a process of kind of collective study and sharing, and it's kind of, the curriculum is going to be open source and available to everybody and kind of keep growing and expanding and keep coming together around various different events after unsound. And so, yeah, I want to, I want to sort of answer the question by saying, I don't know the answers yet, but to get together with people and just begin to pose that question and orient Machine Listening as an object of political concern. And I hope in a few years' time that we can have got a little way down that path and then be able to give you a more satisfactory answer. So could constituting a community, uh, and, and a set of questions to begin with. 
+
+Jasmine Guffond (00:36:51) - And if people want to take part in your unsound Machine Listening curriculum, what would they have to do? 
+
+James Parker (00:36:58) - There's all sorts of different phases of the launch, but if you go onto the unsound website to begin with unsound.pl you'll find some information about it there. And likewise, if you go to liquid architecture.org.edu, because the project is kind of a, it's a, co-production really of, of unsound and liquid architecture. So you can find information about it on either website and eventually we'll have a, a separate website up at Machine Listening, dot exposed, um, that if you're Listening after October, you can, you can, uh, go and check out live. 
+
+Jasmine Guffond (00:37:30) - Yeah. Thank you. Um, this is my first show, so perhaps I can put the link to the Machine Listening curriculum on unsound. I'll ask nudes radio if that's possible, but yeah, 
+
+James Parker (00:37:41) - There's also a Facebook page, although I don't know if I'm allowed to promote, promote yeah. To promote Facebook and the context of what we've just been discussing. 
+
+Jasmine Guffond (00:59:41) - Thanks for tuning in to computational Listening. This was the first of three radio shows as part of my residency at nudes radio. The next additional feature and interview is sound artist, Helen Hesse, and we'll be aired at the same time. On the 19th of October. I'm learning.

content/transcript/smartwife.md

@@ -0,0 +1,128 @@
+---
+title: "Yolande Strengers and Jenny Kennedy"
+status: "Auto-transcribed by reduct.video"
+---
+
+James Parker (00:00:00) - Okay, thanks so much, um, for taking the time to speak with us today. I mean, I just wondered if maybe you felt like introducing yourselves in whatever terms, um, feel right. 
+
+Yolande Strengers (00:00:11) - So I'm Yolande Strengers. I, I would, I would describe myself as a digital sociologist. I've sort of pivoted away from the cord social science disciplines in recent years and have now embedded myself and the faculty of it, and very much sort of working in human computer interaction, design space as well. And so I consider myself quite interdisciplinary as a, as a scholar. 
+
+Jenny Kennedy (00:00:39) - I am Jenny Kennedy and I'm a research fellow RMI T I am in the school of media and communications. I would call myself a major communication scholar. I, um, have a lot, I don't know how to, I always get it's weird, isn't it. I always get really confused on how to describe myself because media and communications doesn't often, it's often really broad. It doesn't say enough about what we are interested in. Um, I think when you say like I'm a digital sociologist, there's something already descriptive in that, in terms of you're interested in the, you know, the sociality of digital devices and media and data and how we engage with them all the time. Whereas media and communications can be, it can be the industry, it can be production, or it can be the type of thing I'm the most interested in, which is how we live with these devices that we have come to, um, just take for granted in our lives and put them under this massive bucket of media and communications from, you know, the, the laptops we browse on the phones we have in our pockets. And now the smart speakers that we bring into our homes as well. Okay.  
+
+James Parker (00:01:58) - And how is it that you ended up working together, um, as a digital sociologist and a media and communications scholar? I mean, and, and specifically on, on this idea of the, uh, you know, writing this book together, um, on this idea of the, the smart wife 
+
+Yolande Strengers (00:02:13) - We met at a conference, um, I was already, I think, aware of Andy's work and we were, um, scheduled to be in the same and the panel together. And our work was very complimentary of both presenting on work that had been a longitudinal study of, of device take ups in the home from different angles. We had different, um, objectives in the projects we were doing, but the work we were presenting was both very much about how the kind of work that goes into operating devices in the home is divided up amongst individuals within the home and finding that that was very gendered. And so we started talking about the, of our work from our interest in that, from our work thinking, this was something we could do to combine the projects and see where it would go. And that very quickly snowballed into this fascination with this feminized, um, agent in the home and your land, you already had a very keen idea to write a book about it. 
+
+Jenny Kennedy (00:03:22) - And I got very easily roped in. 
+
+Yolande Strengers (00:03:25) - I think I was required to write a book about it because, um, the research I was doing was part of my DECRA and on the smart home. And, um, you know, as part of, as part of that project, I had a grade to write a book. So, but the book I thought I was gonna write was not a book that Jenny and I ended up writing, um, because that project was actually more about sustainability and energy effects and smart technologies, but the gender angle was just so big. And, you know, it wasn't something that either Jenny or I had expected. So it wasn't the focus about, about projects, which made it even more striking to, you know, for it to kind of stand up without a save and having been looking for it. Um, and then to kind of consider the broader implications of this gendered nigga coming into the home and also all these different gender interactions with these agents. And there was so much to explore and it seemed so significant at the time that it wasn't being discussed at the level and the depth that it really should pay because you know, these devices are ubiquitous. Now the uptake of them, particularly digital voice assistance, as you would know, I think it has, or is about to surpass the smartphone, which is, which is really, you know, and that is quite significant when you think about the number of days feminized devices and, you know, coming into our homes, the scale and the pace without. Critical attention to what they're doing and how they're impacting and affecting their lives. 
+
+Joel Stern (00:04:59) - Um, could, could you sort of talk about the moment where you decided to call this feminized, divorce devices, smart wives, it, you know, and, and how you arrived at that, um, formulation and, and, and, and maybe just sort of, you know, introduce the idea of what is a smart wife, um, w w where does she come from, and, and what does she do? Well, 
+
+Yolande Strengers (00:05:21) - We're not the first people to come up with a turn there have been other, uh, digital scholars and, um, media scholars who are associated with smart home with a type of wife figure. And in that smart home research that I was working on with another colleague, Melissa Nichols, we were talking about these devices as a type of wife replacement. And so that's not why it was sort of a logical extension from that, but a much broader extension, because the way Jenny and I started to conceptualize it, it wasn't just related to the digital voice assistant and some of the other kinds of devices to be cool appliances in the home that were doing those kind of agenda labor, the vacuuming, or the, you know, remembering the shopping list on the fridge from the smart fridge, that kind of thing. We sort of took it out a lot further and started to think about how pretty much any device you can think of that's coming into the home robotic or smart is taking up these traditional and stereotypical white relaters. And that's why we sort of decided to use this broad category of the term smartwatch to, to refer to all of this, this broad collection across a number of different spectrums and a number of different roles of what a traditional wife was expected to do,  
+
+Sean Dockray (00:06:36) - Maybe it's worth, um, actually recapitulating some of those, those ideas of what the traditional wife means and what the traditional one feels expected to do. 
+
+Jenny Kennedy (00:06:47) - So, um, I guess, I mean, we did a fair amount of research into the history of the wife that I'm sure you're familiar when you, when you start on a project, you're not entirely sure what's going to actually end up in the end product or project. And, you know, the, the history of the wife is, is fascinating in and of itself in terms of, um, you know, uh, an object of patriarchy of mal possession without autonomy often, um, without rights. And so we were looking at that very long history of, um, of the wife and its problematic, um, political history alongside all these, the archetype or ideals of the wife, and especially the ones that came to prominence in the fifties, the 1950s housewife as being one who, um, the home was her ideal domain. The home was also now this, it was this privatized space before the home had often been a site of both privacy and production, but then the home had become very much about, about the private space and the home, the source of the nuclear nuclear family. 
+
+Jenny Kennedy (00:08:09) - And so that was where the wife was located and it was her role to maintain the home, but also to provide any form of care to the family and especially to her husband, the ways in which this 1950s housewife was portrayed was just so beguiling often. And to the point where she's never actually left popular consciousness, and we've even found that this ideal 1950s housewife is not only still present in our popular media, in our TV shows and our movies, but she also comes up in discussions from creators of smart home technologies and robotics. We just can't get away from this ideal that I don't think anyone ever truly lived up to.  
+
+Yolande Strengers (00:09:04) - to. I just wanted to add to that. And then the flip side of that in terms of the why's, we were exploring the book is the contemporary wife in gender progressive societies. And we were very much inspired by it and about Kreb's work on the wife drowns and, uh, how, uh, women in, you know, in Australia and in other, you know, similar kind of kinds of countries have, um, one wives, you know, because the wife is such a critic has been, and has continued to play such a critical role in holding the whole family together. And.
+
+Yolande Strengers (00:09:39) - And so the smart wife is I guess, a market opportunity to put it functionally in a sense that she's responding to this, this drought in contemporary societies, on this woman that was available to do all the things that Jenny just said. And obviously watch still exists, but they don't typically have the time or ability to do all those roles that they should have came and performed, which provides an opportunity for technology to step in and potentially do some of that, that work for us and other things, just to mention that the wife is that we did also structure the book around the roles of the wife. So, um, we started, you picked up on this, but we, we go through four domains of widely labors in the home. So we look at housekeeping, um, caring and emotional labor. And then we looked at homemaking and finally sexual labor. And so we kind of actually look at all the technologies that are stepping in to the home, to do one or more of those things as well, which is another kind of way in which we bring the life into our discussion.  
+
+Jenny Kennedy (00:10:41) - And that's one other thing to add to that as well, which is the history of the wife that is most dominant, is a white, middle-class why the history of nonwhite non-metal class wives is often very much missing from our social histories. And that's one of the, the other, I guess, aspects of the smart, why is this idea of who gets to one who gets to have a wife, but who gets to have a smart wife, there are also inequalities in terms of the types of households that typically have the disposable income to incorporate these technologies into their homes,  
+
+James Parker (00:11:21) - Is the smart life, a technology or an idea or an ideology, or, you know, because I'm thinking about the, the, the really prominent role of cultural representations of the smart wife and your book, um, you know, from, uh, I can't remember the name of the one in the Jetsons, but, um, Rosie. Yeah. They, you know, sort of really iconic, um, I don't know, even if it's an early portrayal, but it is a lot earlier than contemporary ones anyway, but you know, in, in so many films and media, you know, there's a fantasy of the smart wife that seems to be driving so much of so many of these products, which are rubbish comparatively, you know, the smart wives that we actually get are not very smart. So, so what kind of a thing is the smart wife? Is it, you know, is it a technology, um, is it, uh, is an ideology, is it some kind of hybrid, um, how would you describe it as an object of analysis or politics? 
+
+Jenny Kennedy (00:12:24) - I like the idea of the Smart Wife being an ideology.  
+
+Yolande Strengers (00:12:30) - I think it was more that she was a, um, an ideal much like the 1950s house wife was an ideal in many representations across culture and also across technology, uh, as evidenced by robots like Rose and the maid, but also in actual real life robots that were designed in her image or to perform roles that she did. So I think she's, she can be also a technology. And certainly we argue that in the book as well. It's not maybe, maybe it's not that she's one thing will be other, but that she's something that the industry and culture is aspiring towards as well as a physical representation coming about and being manifested in various ways.  
+
+James Parker (00:13:14) - So I wonder if we could direct the conversation a little bit towards digital voice assistance, smart speakers and so on, you know, that's, that's, you know, the, the motivation on our end for wanting to speak with you. I mean, on one level I'm cautious about kind of, you know, emphasizing, listening and voice too much because, you know, Listening and voice are always embedded in other histories, you know, so that the history of a digital voice assistant is not just in the voice and then the, you know, the Listening. So that's really important, but I wonder if you could say something about, you know, what happens with the turn to the voice user interface, like, um, you know, could, could you maybe say something specific about the gender politics of digital voice assistants, like Siri and Alexa and others? And could you say, is it that the digital voice assistant sort of is a key moment in the history of the smart wife or that it, you know, proliferates it, or is it just one amongst many others and many other figures? Um, you know, how important, uh, is the digital voice assistant? And can you say something about it specific agenda politics? 
+
+Jenny Kennedy (00:14:25) - I think the, the introduction of digital voice assistance is a significant moment in terms of human machine interaction, in the sense of.The Swift uptake of these assistants, especially through devices that people were already familiar with. So the introduction of Siri through an iPhone that already has significant uptake, paves the way for other forms of voice assistance to come into the home, then you have the decision made by, um, by the, um, the organizations that produce these assistance to code them in ways that they are going to be most recognizable as female. And I think that that is, I mean, we talk also in the book about why, why that those decisions might be made around our, um, how we have been socialized to expect a supportive voice, to be female, how we're more comfortable with female voices, but it's still, there was still a decision being made to make these assistants, have a female voice and or a female name. 
+
+Yolande Strengers (00:15:43) - One of the things I wanted to add there was that what I think is really pivotable about the voice assistant is it's its role in proliferating, uh, not only itself, so not only digital voice systems, but acting as a gateway technology for voice enabled devices and robots of many different forms and feminizing them. Uh, so, you know, we're seeing now the voice of Alexa and Google home embedded in products that extend beyond their own range, their own, their own cylinders, and, you know, empowering or being used to power, a variety of other appliances and robots in the home as well. But also the technology that has enabled voice assistance is being now embedded into other forms of assistance or other forms of robots. Like the sex robots that we explore in the book has actually a lot of similarities in the, the voice, um, software and the voice activation. That's now being included in these other types of feminized devices as well. So they seem to be this kind of linchpin in bringing about a mass feminization of so many different types of AI and robotics, not just can, you know, contained within their specific kind of niche markets where they started. 
+
+James Parker (00:17:00) - Okay. Could I ask sort of a follow-up on, on this idea of feminization in this context, you know, is the feminization the labor or the vocal character of the device, like, you know, obviously there's been pushes towards, um, gender neutral voice assistance or, or just having the ability to choose a voice of your own, but is that what's really, I mean, how, you know, how much of a part of the story is the voice of the voice assistant in terms of feminization? 
+
+Jenny Kennedy (00:17:30) - It's just the voice, the feminized voice is just one aspect or layer of the feminization of these kinds of voice assistance. The other significant aspect that has been feminized is the types of care and the types of work these devices are put towards doing, um, because they are often performing forms of care and roles that are intrinsically linked to, uh, idealized as feminized in society. Um, but also, uh, typically undervalued as well. 
+
+Yolande Strengers (00:18:05) - And then the other aspect to it as well, is there a feminized personalities, as Jenny saying, you know, it cuts across so many layers, it's definitely goes far beyond the voice. And actually that's something we address extensively in the book is while we're sympathetic to making these devices, gender neutral and to, you know, removing, you know, changing the voice or switching the gender of the voice, we don't think that completely, completely solves a problem because of the roles these devices are intended to perform. And also because of the feminized personalities that they have, we don't see a diversity of feminine attributes or feminine personalities on display. We see a very uniform type of femininity being put on show here. And that's another thing that concerns us because it's the kind of femininity that is compliant that is friendly, that is likable, that is, you know, ready to serve. And please, and of course there's many other types of femininity and many would argue is where you do that. That's actually quite an outdated idea of what women, how women should be uniformly behaving in society in today's age. So there are many problems with the devices that extend beyond the voice. 
+
+Sean Dockray (00:19:18) - Um, I don't know if it's too, too much of a digression, but, uh, in the way that you were describing this history of the life leading into the Smart Wife, uh, and it reminded me of this kind of, you know, informal longstanding kind of, uh, balance of power or agreement between corporations in the state where, you know, in order to reproduce labor power, uh, the wife and the home become the kind of factory for, uh, reproducing workers. And, um, and you kind of hinted at that with the, with the acknowledgement that, um, you know, women, uh, have not been in re Muna rated for that work that they've historically done, particularly in that period from the fifties. Onwards. And in the smart wife in these devices, sort of taking on some of these care giving little aspects of, of caregiving, I guess, that that shifts the relationship. But one thing it does is it turns, uh, this uncompensated work into a service that we pay for. And I guess what I'm wondering is what happens to, we still need that same work to happen, right? People still there's still care. That's given that's that actually helps in like, uh, allowing us to live and have an emotional life and all of these kinds of things. And I don't think these devices are actually providing, even if they're stepping into that role. Um, and so what's happening to that, like, um, relationship between, I don't know, like what, what's the division of labor or something that's happening in the household now between the, these devices, these smart wives and the actual, um, uh, reproduction of labor power that happens in the home? Um, 
+
+Jenny Kennedy (00:21:03) - No, I think, I think, I think you're raising a really important point, which is how these device devices are part of larger capitalist structures that we bring into our homes. And there are also already divisions of labor within the home that are problematic and bringing these devices in as a presumed, um, solution to the conflicts around division of labor in the home. And they're not, they're not being able to live up to that ideal. And mostly it's because they're attempting to displace that labor rather than equalize it amongst the members of the household. And what often at the moment happens is Annabel Crabb, the book, um, inland you mentioned earlier, um, as Annabel Crabb talks about is that most of that care labor falls to women in the house when they're in a heterosexual household. So by bringing devices in, it's not actually, it's not necessarily meaning that for the men and women in the home that they now perform an equal amount of labor. 
+
+Jenny Kennedy (00:22:12) - What often happens is, and this is, this is what came out of the research. And Andy and I were doing that brought us together. Was it often when the, when devices come into the home, one person in the household takes on the kind of curating managing, overseeing role becomes the tech expert. If they're not already. And they are the ones who set up the devices who maintain the devices. And often, especially when you have a complex household that the operation of this interoperable system takes a fair amount of work. So instead what you end up having, having is instead of the care labors that are necessary to be done in the home being divided amongst everybody, you now have an additional set of labors required in the home. That is about the care of these devices. And it's still not addressing the gap in terms of the intimate care labors that are required. 
+
+James Parker (00:23:11) - I just, I've just got this, um, this, this figure of like the Hi-Fi guy in my head, you know, as I'm listening to you. And then that suddenly makes me think, Oh, a smart speaker. It's another kind of Hi-Fi for Hi-Fi guy. 
+
+Jenny Kennedy (00:23:29) - That's it. That's it. 
+
+Yolande Strengers (00:23:29) - What Jenny didn't say there didn't say there was that the digital labour that comes in with smart home technologies and networked devices, in our research at least and other studies as well, it’s more commonly falling to men, and that’s interesting as well. It’s actually quite an invisible labour in the sense that no stats on housework that we could find track this labour. So we commonly know that women do the majority of typical housework in a heterosexual couple. But there’s this new form of labour that’s coming in that could potentially change that. One of the things we were concerned about there was that if men are taking up more time taking care of devices in the home what does that mean for other labour in the home that needs to get done? That has traditionally fallen to women. Again, you haven’t necessarily reduced the labour with the smart wife. You’ve potentially created more and changed the dynamic again in the home and possibly not in a more equal way. So there’s a whole lot of labour politics there in terms of what these devices are meant to solve and what they may actually end up doing.
+
+Jenny Kennedy (00:24:43) - Yeah, you've created another excuse for not doing the washing 
+
+Joel Stern (00:24:46) - Anecdotally. I can confirm that that has happened precisely in my household and we're trying to work against it. But, um, the amount of time that I've spent recently, um, caring for the maintenance of the devices that I've brought into the home.Which sort of stopped functioning and then kind of have to be repaired and, and, and sort of, yeah, it's a, it's a disaster and has produced a lot more unnecessary labor and not necessarily equalized. It makes me think of another figure in your book. Um, resource man. And I was wondering actually, because the voice assistant in your book is, um, where you sort of pivot towards, you know, Amazon and the destruction of planetary systems, you know, capitalistic activism and so on. And you have a whole section in the book there on eco-feminism and the relationship between that discourse and, and the smart wife. So perhaps if there's a way of drawing out, you know, link linking those stories together, um, that would be great. 
+
+Yolande Strengers (00:25:54) - Oh yeah. Resource man is, is similar to that digital housekeeping we were talking about in the sense that, uh, and this comes from my energy work with other colleagues. And what we were finding in that work is that there's, again, typically one person and typically a man in heterosexual households who takes up the labor of sort of managing energy systems. If they have, you know, solar panels or, uh, energy feedback and monitoring systems or batteries, or, you know, kind of some of the more high-tech energy technology, that's now coming into homes, automated system and stuff like that. But their efforts were often in vain to try and save energy or use it more efficiently, uh, because you know, other members of the household more or less did what they want when they wanted. And it was, I guess, the story of resource, man, it's one of caution that, you know, just bringing smart technologies into the home doesn't guarantee that you're going to save energy. 
+
+Yolande Strengers (00:26:45) - And in fact, uh, in some cases it increases energy because these devices require energy themselves or, um, some automated systems, you know, allow you to do more than just save energy. They provide new conveniences and new, new lifestyle opportunities as well, that actually increased consumption. Uh, and I guess that's part of a bigger story of a smart wife in that she is often entangled in these capitalist systems and the, the face or the voice of, you know, companies, huge, huge companies like e-commerce companies like Amazon, whose intention is to embed us in their markets and to sell people products. And most of the algorithms that these devices have are oriented towards those objectives. I mean, Google's objectives are slightly different. They're about, you know, they've data company, but Amazon is, is very much about e-commerce. Uh, and so, you know, the femininity there serves another purpose, the femininity of the devices that is, which is to, uh, make us comfortable with having a major corporation, you, our homes suggesting things that we should buy and should do. 
+
+Yolande Strengers (00:27:58) - And obviously it's, it's a relatively effective strategy. Maybe some would say a very effective strategy, uh, but not necessarily good for the planet or for many marginalized people. And, and this was also a turning point for our arguments around gender equality in the book, because, you know, Jenny and I were quite confronted with this, what we had, what we came to realize was that arguing for sort of more feminist smart wife may only serve the interests of, of white feminists and, you know, women like ourselves, for example, whereas there are many other women and other marginalized people in the world who are sort of entangled up in these systems of labor and environmental extraction and environmental waste that these devices depend on whose lives are unlikely to be served in any positive way by these systems expanding across the globe. So, um, yeah, it was, uh, a moment where we, we questioned whether the SmartLife was actually a good idea for anyone at all. And 
+
+James Parker (00:28:59) - How did you answer that question? 
+
+Yolande Strengers (00:29:02) - Well, we, we, we said that it wasn't a good idea, uh, but that we recognize that, you know, two people writing a book about it was probably not going to stop, you know, five or six, or probably more like 10 or 20 companies around the world from making these devices all from the millions of consumers from buying them. So we, uh, we acknowledged that it was not a good system as it currently stands. And then the remainder of the book is really focused on what we can do to improve it. 
+
+James Parker (00:29:35) - It feels like that's a stance we sort of increasingly have to take. I mean, in our own thinking around Machine Listening more broadly, you know, it's a double bind where you have to say on some level, you know, I'm an abolitionist or, uh, or a lot of have a certain, you know, more positive Stripe than is often represented, but that just can't be the, on the frontier of political action.
+
+James Parker (00:30:00) - I mean, I feel like it's possibly a little bit too early to get to the, sort of the normative project of the book. Um, but you, you are quite explicit about it, you know? So, um, the idea of producing a feminist smart wife is one of the horizons of the book, but you, you have like a whole number of sort of ways forward or routes forward out of this double bind that you, you mentioned. I wonder if it's just, um, as a neat segue worth getting into some of them. 
+
+Yolande Strengers (00:30:30) - Yeah. So we have, we have nine proposals we end on in our 
+
+James Parker (00:30:34) - Not 10? It makes it fell more authentic, you know, if it's nine.  
+
+Yolande Strengers (00:30:46) - Time for one more down the track. Um, and it's interesting that you said they were quite specific. I mean, yes they are, but there are also proposals that are things to build on. They're not like, you know, you must do this and you must do that. They're ideas they're, um, they're meant to be inspirational. And they're meant to sort of get people thinking about different angles of how we could sort of approach and explore this and, and improve the current situation. Yeah. And, and they, they, they all explore quite different aspects, really. Some of the, about the design and how we, you know, we created a smart wife, how we design a feminist smart life and others are about the industry and, and what we do to get more women into coding and it, but also how we actually change what the it disciplines are. And to bring more social sciences into the design of artificial intelligence and think about them as social projects and social designs, not just technical ones. And then others are about how the devices are represented in the media and how they're also developed in science-fiction and popular culture and how those provide inspiration for the roboticists and, and the versions that we see in our home. So we, we, I think they're quite far ranging in terms of the, the areas that we go to and how we can improve the current situation. 
+
+James Parker (00:32:03) - Could you, could you say, for instance, what it means, what it might mean to queer the smart wife 
+
+Jenny Kennedy (00:32:08) - To queer the smart life we're talking about trying to get away from the very narrow idea of femininity that the smart wife currently is portraying to them. Um, and to think more about what other forms of feminine too, there might be. So this idea of the smart wife, always being softly spoken or polite or subservient, maybe she can be a little bit more boisterous or, um, affirmative and can do so in ways that are still positively feminine. Whereas often what we come to do is attribute certain, um, associate certain attributes, either negatively or positively with femininity. This also ties into this idea of, we need to see smart wives perhaps in like in popular culture or in devices that are the, just giving us more range. And that are not part of the current heterosexual construct of their subservient, 1950s housewife 
+
+Yolande Strengers (00:33:24) - And queering. We do talk about currying femininity mainly, but we also talk about querying more broadly in terms of, you know, all genders, you know, potentially being part of this, this story. And, you know, and there is a part of the book where we look at a number of, uh, feminine robots, boy bots, and also kind of more cartoon inspired or animal inspired robots as well. But again, it quite similar in the cuteness and in their, um, the form of femininity or FMI femininity that they, they portray. And so even there, we, we think there's an opportunity to queer, not just the forms that we're seeing, but also yeah, the personalities. And it doesn't mean that they all have to be evil and destructive and rude. Uh, there are there, you know, there are many different types of people that we interact with in our lives. Why is there only one type of personality repeated over and over again, and the types of technologies we interact with? 
+
+Sean Dockray (00:34:30) - Hmm. That's a great question. Earlier, when you were, when you were kind of talking about the, the, the attempts to like, sort of cement into our relationship with, with Amazon, I was thinking about the ways in which the devices kind of prepare the young generation to enter into the market and that relationship with these big services and some thinking obviously of children. And since we're like at this point in the conversation where we're thinking about ways forward, I just wanted to ask you to talk a little bit about the, this relationship between children and the devices. You sort of mentioned that in the promotional material, you know, that they're often shown answering children questions and.
+
+Sean Dockray (00:35:08) - Yeah, that's clearly how they're, how they're sort of used. But I was wondering if you could both talk about the relationship that children have to these devices and also imagine or reflect on whether actually some of these strategies that you're talking about about possible ways forwards might be particularly relevant for, for, um, kids who are, I think are quite open to other ways of thinking and being, and living. So 
+
+Jenny Kennedy (00:35:34) - I think one of the, one of the things around the idea of children using these devices is that children are very open to the possibilities of these devices. They are quick to assimilate them into their day to day lives and they do so from the perspective they have as true children, which is, this is, this is kind of a toy or something fun or something I can engage with, but what they don't have is the critical thinking around what type of ecosystem this device is part of and how their way of kind of engaging with it is actually forming a data profile on themselves that they are not yet able to take ownership of. So children engaging with, um, with devices, listening and learning to sleep to them. Um, so today my daughter managed to tell Google how to turn the TV on. I haven't explicitly told her the command on how to get the TV turned on, but she's been listening to me, engaging with Google and over time practicing, I was encouraging her to go outside and play and she stormed off inside say, I want to watch tele. 
+
+Jenny Kennedy (00:36:53) - And I stayed outside thinking, this is fine, cause she won't fail to turn it on. I've got my phone with me, but I came in 15 minutes later to her sitting very smugly on the sofa because she had to given the command to Google and it had finally responded. So there's a quick to become. Children are quick to become familiar with these devices in their home when they see adults engaging with them to, and see them as being potentially useful. But they're not fully aware of what they're engaging with. And also they're not yet familiar with what the limitations of what they're engaging with. Uh, so they, you know, understanding that how it's connected into perhaps different devices or how the Google, your, or Alexa that you're speaking to on your smart device is the same Google that you using on your smartphone. For example, 
+
+Yolande Strengers (00:37:54) - That's also something that we explore in the book is the research that shows how, uh, children are, um, less likely to be able to distinguish between a computer that's like a human and a real human. So that was another concern of ours is, or is another concern of ours is, is how, you know, if children are again only seeing this one form of femininity and early and are interacting with it on a regular basis just as they are with other members of their household, what effect is that going to have on them in terms of how they save women? Because you know, most of the engaging with these devices, with the feminized feminized voice, unless you actively choose to change it on some devices and you want that, what is that going to mean for the way that they then interact with, with other people? Uh, because if they're not making that distinction between machine and human, then it's all kind of one in the same learning process. Uh, so that's, um, another issue that we raised 
+
+Jenny Kennedy (00:38:55) - And children also, uh, they're seeing that the device will do what you've asked it to do, regardless of how politely you ask it, which is another crucial aspect of that. Not being able to differentiate between words and they're talking to a device or to a human they're not learning the repercussions of inappropriate social interactions. 
+
+James Parker (00:39:19) - Yeah. Well, when my son tries to, he typically wants to get the phone to show him pictures of poor patrol. You just escalate, like when it doesn't understand, he just starts shouting at it. I mean, that's probably, he probably gets that from other places too, but actually, but it makes me think of, so one of the things that, you know, we're interested in in terms of the politics and Machine Listening is both the politics of when Machine Listening works and the politics of when it doesn't work. And that's, uh, you know, there's a whole chapter in your book about this theme. You know, one of the problems with smart wives is.
+
+James Parker (00:39:55) - That they work too well, and they offer too seductive a, a figure of, of domestic labor that, you know, normalizes and entrenches a sort of a regressive four, uh, uh, idea of, of womanhood, um, or domestic labor. But another problem is that they stuff up all the time and it's not just, um, um, miscomprehension but other things too, in, in, in, in the book, you talk about provocatively bitches with glitches. And I just wondered if you could elaborate a little bit on that idea, what you think the politics are of the failure of the smart way. 
+
+Jenny Kennedy (00:40:33) - I think we're going to get t-shirts printed with the slogan bitches with glitches. It seems to be a very popular title. Um, capitalism, why not? Um, I think the bitches with glitches, I think it brings together so many of the things that we're talking about about this very limited idea of how feminized this feminized idea of the smart wife should be behaving and how quickly we are to turn when it doesn't do what we expect it to do, but also the way in which then we discuss that or talk about that. So it becomes very gendered the language. We don't talk about a technology failing. We talk about a feminine entity failing and use gendered descriptive terms as to how that technology has faded. So there are, you know, one of the examples we talk about is a robotic device Cleo, who is an LG device, who has a malfunction during a, um, a tech show.  
+
+Jenny Kennedy (00:41:41) - And instead of saying, you know, there was a malfunction and the device didn't respond, the device is it said that the device was giving the male presenter, the silent treatment as the, you know, this kind of idea of, you know, this, the wicked femininity that it's not the technology. That was the problem. It's the very kind of feminine essence of the technology. That was the problem. So that's one aspect of bitches with glitches is the way in which any, um, limitations of the technology is reframed as kind of this wicked femininity. But then the other aspect of the bitches with glitches is the way in which these technologies can provoke Berry gendered violence, basically in terms of abusive language, insults sexual comments, because they're subservient personalities, but are not able to stand up against any kind of abuse. 
+
+Yolande Strengers (00:42:46) - We liken that to, um, the everyday sexism movement and other sort of similar campaigns that have called out, you know, uh, abuse and small infringements as Laura Bates, I think calls it, uh, that, uh, that are gendered and, and directed at women and, and, and make the link, uh, in terms of not obviously in terms of direct abuse to women, because obviously it's a piece is directed towards devices, but certainly feeding into that culture, that everyday sexism culture, which is, is very much gendered and questioning what impact that might have then on broader society where, you know, on mass or going around and providing gendered comments that aren't very appropriate or friendly to our feminized devices. We're also concerned about the irony here, really because the industries that these technologies are coming out from all of their male dominated industries, you know, particularly the frontiers of technology, we see, you know, the highest numbers of men. 
+
+Yolande Strengers (00:43:50) - So it's actually men designing and programming these technologies mainly on the whole, and yet when they go wrong, we're blaming the feminized technology, which is Eddie's quite all really, but also kind of quite unfair when you think about it, because women are essentially getting blamed for something that is not, you know, the responsibility of that gender. And then again, sort of is feeding into this culture of women being slightly imperfect and glitchy, whereas kind of diverting the attention away from the people who are designing and making these technologies. We also talk about how, uh, in the media, this is also, this also happens and we talk about headlines that actually, uh, blame the devices and, and, you know, it has comic value and it's quite entertaining to say, Oh, Siri needs to wash your mouth out with soap and water, because, you know, she's says all this racist, dirty stuff, but in a way that again, just reinforces our point. That is sort of blaming these feminized.
+
+Yolande Strengers (00:44:50) - Entity or device rather than saying, well, the programmers who, you know, have created algorithms that are racist and sexist, they need to have a hard look at the technology they're designing. So it extends beyond the interactions in the home. It's also about the ways that we're discussing these devices in the media and in popular culture as well. 
+
+Joel Stern (00:45:11) - I was really glad that you proposed, um, sort of feminist coding and, and, um, you know, women programmers as one of the reboots as one of the sort of necessary rebids and, and actually, um, girls who code that they are located in the same building as, as liquid architecture where I work in, in Collingwood at Collingwood art precinct. Um, I was sort of wondering if I could ask him a more sort of speculative question about where you sort of see this going in the coming years, whether you're kind of optimistic that these kinds of feminists re re boot of the voice assistant is possible. And, um, you know, and also with relation to coding, uh, a voice assistant that has been programmed by girls, by women, for women or for you, you know, but what would it sound like? What would it do? And, you know, the kind of, I think that we, uh, um, likely to see, um, that kind of thing happens soon 
+
+Yolande Strengers (00:46:15) - Is actually already a lot of inspirational work happening in this space in terms of how we might design more feminists or just better devices and smart wives for our homes. And there's already, you know, a move in the industry to hire anthropologists and sociologists and other social scientists and bring them into the design phase and, and really start to think about gender issues. So there is positive change happening, I think, but unfortunately, a lot of it's also quite superficial and, you know, it's, it's about sort of just like, we've been talking about neutralizing the voice or changing the voice and saying, okay, we fixed the problem now. And so, you know, let's not worry about the gender question anymore, whereas the issues we're raising in the book go much further than that. And in terms of looking at the personality and the devices, we're bringing into homes and also the roles that they're intended to perform and how that in turn, then changes and displaces and disrupts the, um, agenda labor that already exists in homes and society. 
+
+Yolande Strengers (00:47:14) - So there are, there are many deeper layers to go is what I would say there. And I think, uh, there are, as I said, a number of people who are starting to do that work and some great examples out there in terms of what it would look like. Well, hopefully it would be really diverse. I mean, that, that's where we didn't put prescriptions in. It was, you know, that's where the whole concept of queering is about broadening and diversifying and moving beyond the very uniform types of products that we currently have on the market and experimenting and being playful and coming up with different personalities and ideas. I don't think it's that we need to design the one ultimate perfect smile. One, if it's actually that we need, if this market's going to exist, you know, we need to diversify it and diversify the personalities and the agendas and, you know, the forms that we're seeing in our homes and also the labor's and the roles that they're intended to perform. And also having a fairly serious think about whether they are a good idea, and if they are going to come in, you know, what other impacts I might be having. So it's quite a far reaching set of questions and considerations, I think in terms of how the industry moves forward here. And, um, that's why we did put forward so many different proposals to kind of look at this from a number of different angles. 
+
+Jenny Kennedy (00:48:31) - I can just say that I fully concur with what your client is saying. Um, I think that we are moving in the right direction and the fact that there is a market for this book is an indication that these are ideas that are starting. We're not the only ones who are looking at the current smart home industry, thinking something has to give here. Something has to change that there needs to be more diversity. And we are seeing initiatives, maybe not to the scale we want to see, but I think the broader social currents happening in society right now indicate that they should get better. And I am an eternal optimist at heart. And I also agree in the, we don't yet know what this looks like. And that's for very good reason, because the hope is that it's not going to be one of the very few options we have available to us right now. There's going to be more than one option and all that optimistic.