Still from Camata by Pierre Huyghe, at Palazzo Grassi, Venice, 2024. Photo by Kevin Walker
In the first session, we discussed this artistic perspective on AI. Present were a couple of us using AI in artistic practice, a couple people looking more at archives local history, and one investigating prosthetics.
The local history angle is that all data is local, starting with us and our localities. But soon it becomes ingested and obfuscated, and the local can become lost. It’s not difficult to see the role of AI in this, and we discussed Ted Chiang’s point that a lot of people’s fears about AI are actually fears about capitalism, and how AI might be used against us for someone else’s profit.
We discussed Pierre Huyghe’s exhibition “Liminal” at the Punta della Dogana in Venice in conjunction with the 2024 Venice Biennale, which a couple of us visited. You can read my mini-review on Instagram. Someone else saw it months later than I did, and we discussed how the AI driving some of the artworks should have matured in those intervening months, so we might have had very different experiences of it. On the other hand, we speculated that perhaps much of the data it collects might be very similar, or that it might (in a roundabout way) collect data generated by itself or another AI, thus might reach some steady-state end-game where the work is truly ‘finished’.
Artists Fabien Giraud and Raphaël Siboni’ are creating with Huyghe The Feral, ‘an AI system designed to create a thousand-year-long film, broadcast live, with the participation of thirty-two generations of human performers…. “The more domesticated a machine system gets, the more ‘feral’ we become.”’
In their interview, one of them states, ‘The only definition of the human I can come up with is that it’s something always in the process of transformation, of de-identifying itself to become itself…. ‘Our subjectivities of late-capitalism have nothing in common with the subjectivities of ancient Mesopotamia. We’re almost a different species.’
This was questioned: Computation used to be an embodied act – counting on our fingers (for me this still holds true!). Now we can ‘know’ things outside of our lived experience because of this. But in fact we’ve offloaded knowledge since stone tablets. How different are we, really, from a couple thousand years ago? We discussed the TV series Altered Carbon, in which people download their minds (somehow) into different bodies. One of us, who works with motion capture, already felt comfortable jumping into different bodies. But the larger question is, What happens when our knowledge becomes externalised or disembodied?
So maybe a better question this raises is, What exactly is a human? Doesn’t it mean to inhabit a single body for a period of time, and then die?
A theme running through the article, and our discussion, is about AI as a witness or observer of humans. We already use recorded evidence (from CCTV for example) in court cases, but an AI witness might not have a face that a jury could see.
Some other points from the article:
‘there is a search for a perspective other than the human, an “outside of us” that allows a knowledge of ourselves that we cannot otherwise access, an experience of the impossible.’
‘We’ve always liked the joke by Godard, who had good jokes. He said that if the telescope was invented to show the infinitely big and the microscope to show the infinitely small, the caméscope, the video camera, is here to show the infinitely average, the infinite middle. How can you pierce the infinite average with the infinite outside?’
Generative AI provides one answer here, as a viewer of all the content on the internet. Regarding the ‘infinite outside’, philosopher Federico Campanga says the artist signifies more of a position than a person: an artist, like a shaman in some cultures, can step outside a reality system to show other possibilities.
The artists discuss what they call ‘the anthropological function of art’ as opposed to art as social commentary: ‘In the big scheme, what is art doing to or for our species?…. I don’t think of art as some kind of activity humans do, but as that through which we become humans in the first place’.
Computation, then, is discussed as a means for prediction, helping us deal with uncertainty.
‘It’s striking that, as we are developing AI, we are destroying the place in which we live, which was, until now, the very substrate of the possible. We might just be creating the witness of our own disappearance. Maybe that’s what AI will do.’ A key point here that we discussed is AI as witness – in this process, but also more generally as an observer of humans and the world. If AI is a witness, we discussed whether it could be a juror.
A starting point for The Feral project ‘is the question of how an AI learns what a world is…it’s not conceptual knowledge. It’s purely statistical, just information….. Artificial intelligences learn from different outcomes. It is a superposition of dimensions that generates infinite possibilities.’
Bold text in this post is by me, in order to highlight some key points.