For Auction: Art Created by Artificial Intelligence from Artists Also Created by Artificial Intelligence

This striking abstract painting is by an up-and-coming artist named Taylor; you can bid on it for $200 or more soon, in hopes it becomes a rare collectors’ item that symbolizes the start of a new art movement.

Because the thing is, “Taylor” is not an artist and this is not a “painting” as we typically understand those terms. The painting was generated by artificial intelligence… as was the painter herself. 

You can see more works by “Taylor” here, part of a fascinating project by rising technologist August Rosedale, whose development I’ve been casually following since the early 2000s.

Starting soon, Rosedale’s Mirage Gallery will auction off AI paintings like this to the highest bidder, and confirm their authenticity by registering them on a blockchain.

“I think that people will bid on AI-made paintings for the same reason people bid on human-made paintings,” August tells me. “People enjoy having beautiful art pieces around them and also can be invested in the idea of the artist's work increasing in value over time.”

The painting, to be sure, is not actually painted, and will come with a blockchain token.

“For the physical pieces, it will be a 24x24 inch canvas print. Each buyer can choose to sell the token and canvas print separately but the value will come from keeping them together.”

To create Taylor’s work, August compiled ten thousand or so abstract art paintings in a wide variety of styles, then fed them into a generative adversarial network, or GAN:

“The GAN works by having something called a discriminator whose job is to determine if an input image was from the original dataset or not. What this means is that the model that is training will create an image, and then feed it into the discriminator and see if it can convince the discriminator that the inputted image was from the original dataset. Over time, the model is able to create more convincing outputs that start to trick the discriminator.”

The only human input in their creation happened at the beginning, when August curated the datasets the AI then worked from:

“It is important for the datasets to have similarities in the images or else the outputs will be much more difficult to interpret,” as he explains. “Once the dataset has been made, the GAN plays a sort of game with itself to try to trick itself into believing that its own creations were from the original dataset. With powerful GPUs and lots of computing time, this process can result in amazing pieces of art. The Mirage Gallery artists will continue to refine their art with more hours of training as time goes on.”

Same goes for the personae and appearance of Taylor and “Alejandro”, the Mirage Gallery’s current representational artist:

“The actual faces of the artists are created with the same type of GAN, but instead of training it on art pieces, it was trained on human faces and can generate faces that are undeniably human.”

August believes AI-driven “artists” like his will compete on the market with actual artists. In fact he believes they will raise sticky legal and philosophical challenges:

“There will need to be legislation put in place that determines ownership of images generated by an AI,” as he puts it. “Should it be the person who created the neural network? Or should it be the people whose images/artwork were used in the training dataset?

Art by "Alejandro"

“One of the frightening differences between AI and human artists is that the AI artists can produce an infinite number of artworks. Each piece can be generated in under a second once the model has been trained. This will end up causing us to live in a world where the majority of art is created by AI and it will be more expensive to own a piece of art created by a human artist. No one fully can comprehend the changes that will happen in this world due to the increasing power of AI.”

That brazenness may open up massive cultural dilemmas, but I suppose they’ll be less risky than, say, letting AI run wild on military targeting systems.

And if the wild breadth of August Rosedale’s vision (not to mention his name) sounds familiar: Yes, August is Philip Rosedale’s son, who I first met when he was a kid wandering around Linden Lab, his head barely visible above the desks. But he tells me his vision for Mirage Gallery is not directly related to his father's virtual world work:

“While my dad always has mentioned the power of AI, this project wasn't inspired by his work with Second Life and High Fidelity,” as he puts it. “Mirage Gallery came from my own interest in different AI architectures and myself just wanting to see what AI is capable of… [and] bring it to the market to allow others to see what is to come in the future of AI.”

Explore the Mirage Gallery here.

Comments

[Joey1058]

There is going to be a time where machine produced 'anything' will require a license, or some kind of generic contract. Perhaps an AI guild. There will be those that will clamor for AI productions, and those that will abhor anything even resembling potential mass production.

If there is any kind of 'robotic uprising', this is how it's going to start. We'll be mesmerized by a soothing orchestral bot opus, while we peruse Art Institute quality paintings.

My question now, is, whose photographs did the AI use to generate Taylor and Alejandro?

[Kitty Revolver]

Humans will continue to create despite the advances in AI and people continue to sell and buy art directly from artists. Will AI’s push any real original or political boundaries? No because it is just fed information and “regurgitates” what it has learned. Can the next Cindy Sherman really be original when it’s just follows the patterns of Cindy Sherman? Is that art? It certainly can mass produce art at a large scale, but what does that mean when you rip images from the internet to create a bot’s original work. These are images and not really art that has any value to the human spirit. It maybe efficient, and art isn’t efficient.

[TroyMc]

One could train an AI to produce works that look like new cubist works by Picasso, but would it sell at the same prices as the real works? Of course not. Picasso is dead, everyone knows that, and the list of his cubist works is known and finite.

Part of the perceived value of an artwork comes from the fact that it was made by a fellow human with a real life story. Their works emerge from, and are part of that story, a story which always ends, because we're all mortal.

Of course, a good writer could create an artist character with interesting life stories, and AI-generated artworks to go with them. And nowadays it's even possible to auto-generate stories, to some extent. To make that work though, I think he writer using the AI would have to keep their use of the AI secret. If the secret ever got out, it would be a big fiasco. But maybe the fiasco itself would become part of the story, which would increase the value of the artworks! I guess the key thing is to keep the number of artworks finite. Anything with an infinite supply and a finite demand will have a low price. That's Microeconomics 101.

I think there are still some business opportunities, just not in the world of fine art. Think clip art, stock photos, or commodity illustrations, i.e. Shutterstock, not Christie's.

Comment by Troy McLuhan in SL.