»genesis - v1«  (2023) [Live @ZKM, 2023]
for no-input AI neural generators

genesis - v1 is a first attempt of using audio generative neural networks without the resort to any human data, amplifying the own materiality of the algorithms towards cybernetic chaos rather than nature imitation. While the use of neural generative modelling for reproduction becomes invasive in our lives, their specific materiality is largely unknown, unshown, and even hidden, as if its revelation would be the unveilement of a magic trick. However, what could happen if we take the opposite position, that is enforcing individuality and emergence away from any human data and judgements? While learning without data can seem non-sense, I tried to overcome this apparent paradox by rather by giving to a set of different agents and set of structural (and often impossible) objectives, to force them to find the genesis of an eco-systemic equilibrium with which I communicate during the performance through diverse inputs (request, bending, recording).

»genesis - v1«

Aletheia is a part of a cycle of audio-visual pieces, Daath, that explores several ways of composing and co-creating with neural audio and image synthesis. This musical piece echoes the concept of aletheia, a pre-socratic allegory of reality later described as an unveilement by Heidegger, by corresponding the underlying version of reality subsumed by machine learning processes, and compositional processes based on exploring, transforming, hacking, bending, and interplaying with models. While every sound present in the piece are generated with neural audio synthesis, Aletheia organizes a game with reality by decomposing and deconstructing sounds from real world (voice, instruments…), illustrating the of the discovery of reality by going from phenomenological pure perception artifacts to an imitation game (corresponding to origin of the word “a-letheia”, extraction from oblivion).  However, the more the generating models are pushed to their limits and submitted to complex manipulations, the more they produce a deranging equivalent of reality, showing the destructive counterparts of the modelling act and the aporetic aspect of extracting the inner nature of phenomenas through observation and quantification. Hence, this piece is a reflexive work between philosophical aspects of reality and exploration of model’s rules through deviation, while being motivated by the development of new approaches to compose and interact with neural generative models and the experimental research of new aesthetics that these novel methods can offer.

Axel Chemla--Romeu-Santos

Popular tracks