#29: Harold Cohen - The AI Artist
This week: Harold Cohen and the AI Artist // Reader Survey ... results are in // The week that AI changed my life // Things to see and to this week in the Creative R&D Ecosystem
This week with my new piece out in The Art Newspaper, I wanted to quickly look at Harold Cohen, the artist who defined what it means to be an AI artist. Plus, how AI’s been changing my own working practise, and the usual things to do links.
Thanks so much for reading.
ART! // The art world’s AI dilemma
Being given the opportunity to write for The Art Newspaper, journal of record for the art world, has been an amazing offshoot of writing this newsletter.
And writing two cover features this year is a crazy privilege I have no real way to compute - first on Immersive Institutions, now on AI, artists and museums.
Please go and read the piece.
But one note on Harold Cohen before you do.
In the article I write about how Harold Cohen first defined what it means to be an “AI Artist” and how that definition is now under threat from a rapidly changing technology.
This is just to re-state how important I think Cohen is - and how much greater our sense of his importance might become as this technology impacts on us.
I’ve been lucky to spend a fair amount of time with his work thanks to the excellent Mila Askarova and her team at London’s Gazelli Art House, brilliant pioneers and defenders of AI art in the commercial gallery sector. Their support for and representation of his work is timely and imaginative - this is just one of various events they’ve ran to help others understand him better.
Cohen is a bridge between artistic tradition and new worlds of digital art. The fact that he made AARON, his AI partner, a painter MATTERS - that these pictures sit in frames on walls in quite normal ways gives them a connecting point between past and future that many brilliant digital artists have ignored - and so left their work prone to technical obsolescence. I talked about this back in Issue 16 and the work of Aureia Harvey - often so hard to find now.
Whatever happens to the role of the AI artist Cohen defined, I hope his work’s brilliance continues to be better understood.
Go find it. And please go read the article and tell me what you think.
IDEAS! // What to watch, read and see in Creative R&D
Melissa Terras is one of the most important academic leaders around questions of culture, creativity and technology, and a thoroughly excellent human to boot. This report she’s led on the Creative Industries and Artificial Intelligence is required reading.
Good interview with body horror film-legend David Cronenberg, one of my total heroes, in Variety, on the promise and threat of AI for film-makers. My take is it allows one vein of film-making to become like the novel, poetry or painting - an individual pursuit rather than collective endeavour. That has huge industrial implications, but also exciting artistic ones - the novelists of film like Bergman or Ozu, or Chris Marker, flourished because of very, very supportive financing. Few others have ever managed that. This could be their chance.
LOVE this - Nigeria has launched its own AI model trained in multiple indigenous languages. This is an area of such rich potential - and we shouldn’t stop at living languages. I think of something like ancient Mesopotamian or other languages kept alive by small cabals of academics and curators. What a chance to bring them back and preserve them.
This is the kind of research I love. Neuroscientists have been using VR to get a better understanding of the brain’s own GPS system, and to start to figure out how “out of body” and other hallucinatory experiences happen as a result. The brain works to a kind of grid cell system, that tells it where it is at any point - when that fails, you get to hallucinations and possibly more. I assume this also has something to do with dream states. We talk so much in the arts about simulating these kind of feelings - understanding how this works might be the chance to create work which actually DOES this.
Also on the VR tip, this early stage research on ways to make headsets radically lighter is a good path to follow. Keeping battering away putting screens in front of peoples eyes locks us into an optical technology problem that’s not solveable fast, and has radically hurt adoption. Proper activity on alternatives is to be thankful for.
READER SURVEY // At least that’s what you said…
Thank you SO much to everyone who answered the reader survey I sent out a couple of weeks ago.
I thought i’d share what you said about the main question I asked - what new kinds of content you’d be interested in for the newsletter?
So the answer is … excellently unclear. More of everything, but don’t mess up what you’ve got is my takeaway.
I will definitely be adding some video interviews in the coming months - more than as a chance to talk rather than write once in a while, and to dust down my video editing skills.
And the live online - and maybe IRL - events piece is definitely on the list.
So, keep ‘em peeled reader, there’s more to come.
TECHNOLOGY // The week AI changed my life
I’ve spent the last few weeks grappling with what AI might mean for artists and museums.
But slowly and subtly over that same time period, AI has crept into my day today and become a permanent feature in a way it hadn’t before.
I’ve been with it since the ChatGPT moment in November 2022. From the endless replace-your-face memes on friends’ WhatsApp groups to the Sysco Creative AI Challenge I led last summer with the Outernet, UCL and Goldsmiths, I’ve got my hands dirty.
I’ve written about it by my count every other week here on the newsletter.
But honestly, for the things I do day today - the different kinds of writing, thinking and analysis I do as a consultant - AI was still a slightly distant thing, not quite solving the real problems I have. A distraction rather than the productivity miracle we’ve been promised.
But that’s been shifting in the last few weeks, and suddenly I got it.
Got what AI was going to do for me and why just at the moment when - slightly unbidden - AI suddenly re-defined the tools I use every day to work.
And now I get it I don’t think there’s any turning back.
A few things happened all at once.
First I started using Claude to help with landscape and market analysis. I do a lot of this - trying to define large scale groups of different institutions or features. And honestly, I’m not sure I’ll be able to charge people for it in the same way because of how GOOD Claude is at it.
Then Google added Gemini, their AI bot, into Google Docs just as ChatGPT 4 was being released.
I live in Google Docs professionally, everything I do starts there. And so Gemini is an intervention at a very intimate level. It’s demand to summarise everything is ANNOYING - I can read this three line email thanks buddy - but it opens up some genuinely helpful new ways of working. Throw it a bunch of messy data and ask it to make a table and it sorts it brilliantly.
But then the big moment came during a workshop on AI I ran for Museum ID (more dates to come - i’ll let you know when).
Asking groups and individuals to build definitions of what AI is and the strategic risks and opportunities it posed, I asked some to work individually and some to work with chatbots as collaborators.
The answers that came back from collaborative human intelligence empowered by AI were so good it REALLY made me hopeful. The best of us plus the best of tech is a good place to get to.
But there were still gaps - AI’s blindspots matching our own in ways that show how we’re going to need to keep forcing ourselves and it to tough thinking.
Looking at risks AI might help with, no-one in the room said environmental change or sustainability … and neither did the AIs.
That’s stuck in my head as a gap in our thinking and a gap that must exist in the training data.
We have to keep looking for those gaps.
But the practical value of the tools is there for me now, and I won’t go back.
Musical references:
“At least that’s what you said” is a song by the brilliant Wilco. Alt-country weirdos, and I think one of the few bands who write songs that sound like the Beatles and very occasionally reach their impossible level.
Here’s them performing it live in 2009.
See you next week.