Forum Vertigo 2020

Interview with Hugues Vinet

Forum Vertigo zooms in on the vertiginous present, introducing a series of international encounters between scientists and artists, engineers and intellectuals. After three editions devoted to the artistic uses of 3D, code and living material, this year’s Forum Vertigo examines the multiple links between artificial intelligence and contemporary creation, on the occasion of the "Neurons, Simulated Intelligence" exhibition (see p 24). Hugues Vinet, Director of Innovation at the Institute for Research and Coordination in Acoustics/Music (IRCAM), and encounter programmer, explains the reasons why artificial intelligence techniques are experiencing such spectacular success in the domain of art, as in other activity sectors, and provides an update on new questions posed by their use.

This fourth edition of Forum Vertigo is devoted to artificial intelligence. Can you tell us more about what's at issue in these encounters?
Hugues Vinet – Artificial intelligence technologies have made spectacular progress in recent years with deep neural networks ushering in a radical paradigmatic change in the processing of digital data, between the usual approaches using modelling and those using big data analysis. We are beginning to glimpse the great potential for artistic creation: totally new processes for generating and hybridizing content, producing artefacts endowed with autonomous behaviour, etc. In addition to renewed forms and materials for creation, the large-scale deployment of artificial intelligence techniques in society also leads to critical positioning in the artistic field. These encounters bring together some of the most cutting-edge artists and researchers in these fields to present their state-of-the-art work and debate the issues in this ongoing revolution.

Will computers be able to become creative and will they thus replace artists, or is that pure fantasy?
HV – We have a long way to go. For me, the distinctive characteristic between machine and human potential is situated between the notions of generativity and creativity: there is a potentially considerable increase in the generative power of machines corresponding to the scope of the learning corpuses used, but in the end of the day, they are still configured by man. The issue for artists, as it always has been in the history of techniques, is thus to appropriate these new methods in the service of their expression: making a work is a creative act that cannot be performed by a machine.

What are the most promising perspectives for AI in scientific research and in the arts in particular?
HV –We still have a limited comprehension of the representations of the world that are specific to deep networks: they are black boxes that produce totally original artefacts although we can't really compare their logic to our cognitive categories. Having a better understanding of these representations, formalising a theory and rendering them accessible for non-computer scientists is a fascinating scientific challenge

Interview with Hugues Vinet

Ircam

In Code couleur n°36, january-april 2020, p. 28-29

Programme

Mercredi 26 février

14h30 : Introduction et présentation de l’exposition «  Neurones, les intelligences simulées  »
15h00 : L’architecture et la formalisation des modèles numériques
17h00 : Limites et potentiels génératifs de l’Intelligence Artificielle
19h00 : Débat : Vers une IA humanisée ?

Jeudi 27 février

11h30 : Impasses et devenir de la cybernétique
14h30 : Perception humaine et artificielle
16h00 : Générativités musicales
17h45 : La danse entre vivant et artificiel