Timeline

Soft Launch; Broadcast Begins  ————————

28 Jan 2022 : 18:00 UTC+1

Radio Silence; Broadcast Ends  ————————

31 Mar 2023 : 00:00 UTC+1

< Home

The Poetics of Prediction and the Autopoietic Archive

Machine Learning Systems as Unstable, Collective Vocal Memory

Artwork #1


In Search of Good Ancestors

Ahnen in Arbeit

2021 —

2022 —

2023 —

(WORKSHOP#4) Remnants of Future Voices

10 Dec 2022 : 12:00-17:00 UTC+1

In this workshop we will explore ancestral cosmology through the figure of the clone in artificial voice technologies. The workshop will experiment with artistic approaches to future and past-oriented dialogues using voice avatars of ourselves and others, considering the implications of creating a clone that may speak to generations beyond our own lifetime and say things we have never said. This workshop is crafted in collaboration with Angeliki Diakrousi of Varia.

Seeking Truth Seeking

31 Mar 2023 : 00:05 UTC+1

We bid farewell to a year long broadcast with a ritual of voices. A six part improvised performance between Jonathan Reus, the Broadcaster, and vocalists Faye Houston, Jo Bramli and Kassia Zermon, this programme will be played on German Public Radio just after midnight on March 30th to coincide with radio silence.

(WORKSHOP#2) An (in)audible chorus

27 May 2022 : 14:00-17:00 UTC+1

In this second workshop we will explore ideas of audibility in relation to voice synthesis and text-to-speech AI systems, to consider the perceptible and imperceptible voicings we might encode for future listeners. Through using our own voices, and collaborating with artificial ones, we will attempt to tease out the features and meanings of voice in the digital domain, and consider the audible sensibilities we may want to attune with, as we co-create a hybrid chorus. This workshop is crafted in collaboration with Eleni Ikoniadou and Varia.

(WORKSHOP#3) Ancestral Voice

22 Sept 2022

A workshop exploring ancestral cosmologies of voice data within the Critical Inquiry Lab of the Design Academy Eindhoven.

(WORKSHOP#1) Re-Reading / Re-Writing

12 Mar 2022 : 14:00-17:00 UTC+1

How would you like an artificial reader to read? What kind of memory is preserved in an algorithm? What knowledge, thoughts, feelings would you want to be a part of such processes, and what previous »machine learnings« could be refused? In collaboration with Varia.

KONTINUUM Radio Feature

6 Feb 2022

Radio feature about the work played on German Public Radio (Deutschlandfunk Kultur) and Austrian Public Radio (ORF Ö1)

In Search of Good Ancestors / Ahnen in Arbeit is a year-long experiment in generative radio. An artificial BroadCaster with a growing number of voices shifts through speech and song, words and soundscapes, as it extends and reconfigures altruist Jonas Salk's 1977 lecture, "Are we being good ancestors?"

The BroadCaster performs using deep learning algorithms that predict future words from past contexts. Beginning with Jonas Salk’s lecture, the BroadCaster attempts to continue the conversation indefinitely. Ideas of long-term thinking, wrapped up in human activities of prediction, risk analysis, and memory happen at many levels of human society and individual consciousness. How do we navigate the long-term future we are a small part of?

 

Please submit your own thoughts about what it means to be a good ancestor and it will be added to the BroadCaster's archive. Throughout the year there will be a series of workshops meant to collectively contribute to and reshape the BroadCaster's words and voices.

 

The BroadCaster began speaking on January 28, 2022 and has been speaking for 428 days 13h 29m 05s

Broadcast Launch Event (CTM FESTIVAL)

6 Feb 2022

In conversation with Eleni Ikoniadou. Together they discuss voices, oracles, future telling and ego death in becoming archived in datasets and machine learning systems.
No media selected. Please add or select a media file for this tool.

Broadcast Recording 12 Apr 2022 : 11:00:02 UTC+1

Shortly after the first Consolidation

Web Portal 13 Mar 2022 : 18:11:27 UTC+1

Two weeks after going live

Web Portal 6 Mar 2023 : 14:26:46 UTC+1

25 days before end of broadcast

News Public Events

 

Mar 31 2023 : 00:05 UTC+1 (and free to listen thereafter)

(Radio Silence) <a target="_blank" href="https://www.hoerspielundfeature.de/seeking-truth-100.html">Seeking Truth Seeking.</a></i></b> We bid farewell to a year long broadcast with a ritual of voices. An improvised performance between the Jonathan Reus, the artificial Broadcaster, and vocalists Faye Houston, Jo Bramli and Kassia Zermo, this programme will be played on German Public Radio just after midnight on March 30th to coincide with radio silence.</div></li>

 

Mar 24 2023 : 12:45 UTC (Mozilla Festival) <a target="_blank" href="https://schedule.mozillafestival.org/session/QTHPSU-1">Many Voices: Towards a shared vocal practice between humans and machines</a>

Together with Eleni Ikoniadou we have a fun and light session around possibilities of rethinking voice through collectivity in artistic practice. The session will include a short discussion of In Search of Good Ancestors, as well as Eleni's album project Future Chorus. The session is curated by Sophia Bazile as part of the AI Cosmologies track of Mozilla Festival.

 

Dec 10 2022 : 12:00-17:00 UTC+1

(Workshop) <a target="_blank" href="http://varia.zone/en/remnants-of-future-voices.html">Remnants of Future Voices.</a>

In this workshop we will explore ancestral cosmology through the figure of the clone in artificial voice technologies. The workshop will experiment with artistic approaches to future and past-oriented dialogues using voice avatars of ourselves and others, considering the implications of creating a clone that may speak to generations beyond our own lifetime and say things we have never said. This workshop is crafted in collaboration with <a target="_blank" href="https://w-i-t-m.net/">Angeliki Diakrousi</a> of <a target="_blank" href="https://varia.zone">Varia</a>.</div>

 

Oct 8 2022

In Search of Good Ancestors / Ahnen in Arbeit published in the <a target="_blank" href="https://dl.acm.org/doi/10.1145/3546155.3547294">Association for Computing Machinery.</a></i></b></div>

 

Sept 22 2022

(Workshop) <a target="_blank" href="https://www.designacademy.nl/p/study-at-dae/masters/critical-inquiry-lab">Ancestral Sounds.</a>

A workshop exploring ancestral cosmologies of voice data within the Critical Inquiry Lab of the Design Academy Eindhoven.</div></li><li>

 

May 27 2022 : 14:00-17:00 UTC+1

(Workshop) <a target="_blank" href="https://www.ctm-festival.de/festival-2022/programme/schedule-part-1/event-detail/event/ctmevent-189">An (in)audible chorus.</a></i></b> In this second workshop we will explore ideas of audibility in relation to voice synthesis and text-to-speech AI systems, to consider the perceptible and imperceptible voicings we might encode for future listeners. Through using our own voices, and collaborating with artificial ones, we will attempt to tease out the features and meanings of voice in the digital domain, and consider the audible sensibilities we may want to attune with, as we co-create a hybrid chorus. This workshop is crafted in collaboration with <a target="_blank" href="https://www.ctm-festival.de/festival-2022/programme/artists/e/eleni-ikoniadou">Eleni Ikoniadou</a> and <a target="_blank" href="https://varia.zone">Varia</a>.</div></li><li>

 

March 12 2022 : 14:00-17:00 UTC+1

(Workshop) <a target="_blank" href="https://www.ctm-festival.de/festival-2022/programme/schedule-part-1/event-detail/event/ctmevent-190">Re-Reading / Re-Writing.</a></i></b> The first workshop in a series of three. How would you like an artificial reader to read? What kind of memory is preserved in an algorithm? What knowledge, thoughts, feelings would you want to be a part of such processes, and what previous »machine learnings« could be refused? In collaboration with <a target="_blank" href="https://varia.zone">Varia</a>.

 

Feb 6 2022 : 00:05 UTC+1

<a target="_blank" href="https://www.hoerspielundfeature.de/klangkunst-100.html">KONTINUUM: Ahnen in Arbeit on Deutschlandfunk Kultur and ORF Ö1 Kunstradio</a>.

Deutschlandfunk Kultur and ORF Ö1 Kunstradio Austria have commissioned Anne Wellmer to create a 1-hour audio piece 'behind the scenes' about <i>In Search of Good Ancestors</i>. The programme mixes interviews, discussions and music into a radio collage about the work as it begins its journey.</div></li>

 

Feb 6 2022 : 18:00-19:30 UTC+1

<a target="_blank" href="https://www.ctm-festival.de/festival-2022/programme/schedule-part-1/event-detail/event/ctmevent-179">Launch event at CTM Festival.</a></i></b> Jonathan Chaim Reus in conversation with sound artist Eleni Ikoniadou. Together they discuss the ideas behind <i>In Search of Good Ancestors</i> - voices, oracles, future telling and ego death in becoming part of a dataset.

 

Credits

<a target="_blank" href="https://jchai.me">Composition, Development and Concept Jonathan Chaim Reus</a>


<a target="_blank" href="https://dxr.zone/">Website Design and Development DXR Zone</a>


<a target="_blank" href="https://varia.zone">Workshop collaborations Joana Chicau, amy pickles, Cristina Cochoir, Angeliki Diakrousi ( VARIA, Rotterdam )</a> and Eleni Ikoniadou


Special Thanks:

<a target="_blank" href="https://www.matangover.com/about">Matan Gover ( VoctroLabs )</a>

<a target="_blank" href="https://victor-shepardson.github.io/">Victor Shepardson</a> <a target="_blank" href="https://iil.is/">( Intelligent Instruments Lab )</a>

<a target="_blank" href="https://github.com/acids-ircam/RAVE">Antoine Caillon ( RAVE / IRCAM )</a>

<a target="_blank" href="https://coqui.ai/">Eren Gölge &amp; Josh Meyer ( Coqui AI )</a>


The work is commissioned by <a target="_blank" href="https://www.deutschlandfunkkultur.de/">Deutschlandfunk Kultur</a>, <a target="_blank" href="https://www.ctm-festival.de/">CTM Festival</a>, and <a target="_blank" href="http://kunstradio.at">ORF Ö1 Kunstradio</a>.

data ecologies and machine

learning models

Long Durational Compositional Approaches

Lotke Volterra

Collective Data Creation Workshops and Performances: Moments of Embodied, Convivial, Consentual High-Context Collaboration (Rather that Datafication -> Address/Relation/Voicing/Archive Building/Gifting)

Originating Compositional System: Single TTS Model, Single Text Model, Basic Script System, Web Portal, Web Portal Text Submission

Ongoing New Functionalities: New Model Architectures (VITS, YourTTS), Multi-model System, Magic Words System, Prompt Context System Using Small Texts and Text Segments from Contributed Texts, Fine Tuned Models using Contributed Texts

SELFIE

The "Self" target voice models include datasets of approximately two hours of recordings of the performer's own voice, recorded over the course of a month in the studio of the Intelligent Instruments Lab in 2022. The aim of this dataset was to capture a wide variety of vocalizations, a "snap shot" of the performer's vocal production across varied techniques both phonetic and paralinguistic.

 

In producing these recordings, the performer prepared a selection of vocal warmup and training exercises, alongside specific scores and phonetically balanced texts to provide comprehensive coverage of diverse vocal sounds, including:

 

  • Vocal Techniques: Vocal warm-ups and technical exercises in blues and gospel singing, death metal growling, and the hollars, shouts and warbles characteristic of Appalachian folk singing.
  • Paralinguistic Expressions: Humming, breath sounds, percussive lip smacking.
  • Speech Sounds: Readings of the International Phonetic Alphabet (IPA) at different speeds, a complete performance of Kurt Schwitters' abstract sound poem Ursonate, and English scripts commonly used in voice research and voice banking such as the Harvard Sentences and the Rainbow Passage.

Evaluation of the resulting model was done subjectively by the performer, based on the model's ability create a real-time voice clone that was as perceptually transparent as possible, meaning that the model could faithfully reproduce speech and nuanced timbral and expressive qualities when the input voice was the performer's own. Beyond transparency, a secondary evaluation metric of the "My / A Snapshot" models was their capacity for being "pushed" through latent space manipulations into producing new vocal gestures not present in the dataset. This metric was chosen so as to create a model that would be capable of creating subtle but uncanny disjunctures between the performer's biological voice and a "not quite right" reproduction.

long-durational generative radio broadcast timeline

systems lutherie

                & instrumental relationships