Christie’s announced that it is going to auction for the first time art that has been generated via an artificial intelligence algorithm. The art was created by Obvious, a Paris-based collective consisting of Hugo Caselles-Dupré, Pierre Fautrel, and Gauthier Vernier using AI software.
Richard Lloyd, Christie’s International head of Prints & Multiples discussed AI art in an interview on CBS:
AI Art Used Software Called the GAN Algorithm
Obvious used a nifty piece of software called the GAN algorithm. What they did is uploaded thousands of images to a computer and that point it actually splits itself in two. One half is called the Generator and that analyzes those thousands of portraits and learns what a portrait is. It sort of parses through all those and then thinks now I am going to start creating my own versions of those.
The second half of the computer, the Discriminator, tries to spot that. Everytime the title is run, if the Discriminator is able to say that a portrait is created by a computer the Generator runs it again. The cycle finishes when the Discriminator says I give up, I can’t tell the difference between the computer generated version and the human-generated version and that’s what pops out.
AI Art to Sell for $10K
We’ve estimated that it will sell for $7,000 to $10,000. We put a lot of thought into the estimate because if we put a huge amount people would say what are you basing that valuation on because this is the first.
But also we thought that it was the right sweet spot where people would respect it as a work of art because the creators certainly think of it as that.
Who is the Creator of AI Art?
Who created the art? Is it the person who wrote the algorithm? Is it a combination of the artwork that was uploaded? Is it the people that tweaked the software? This is why this is inspiring and interesting because we’ve never really had to ask those questions before.
I remember reading years ago that when TV news started, print journalists thought well that’s it, who’s going to read a newspaper? Everybody is just going to watch it. But in fact, both exist side by side. So I think that in the future in five to fifteen years time there will be parallel tracks. There will be human art and artificial intelligence art.
AI Human Hybrid Art is the Future
There will also be a hybrid which I think is coming down the pike in the near future. Artists have always been great early adopters. Warhol adapted screen printing which came from commercial packaging.
Photographers took the camera and thought we can do weird and wonderful things with this. I think human artists will be working side by side with this algorithm to create hybrid art. It’s just the beginning and is so fascinating in what is going to be created.
Just Don’t Call AI Art a Masterpiece
I’ve done a lot of research into AI art and there is something about using human-centric words like masterpiece. You just kind of stop short. I think a great work of art is a link to another person. You think of Vango and what he was going through to create that. But this is an algorithm so…