Ai-Da the art robot: art genius or twisted male fantasy?

Until 7th November, Sotheby’s New York is auctioning off a painting, A.I. God. Portrait of Alan Turing (2024) by Ai-Da, “the world’s first Ultra-Realistic AI humanoid robot artist.” Sotheby’s says it’s making history with the house’s first auction of “an artwork made by a humanoid robot using artificial intelligence algorithms.” If you believe Ai-Da’s creator Aidan (get it?) Meller, this work questions “what it means to have an increasingly blurred boundary between the inorganic and organic, the human and the machine.” If you listen to Ai-Da, “my artwork uses a fractured and multilayered approach, and this shows the deeper emotional and intellectual layers of Alan Turing himself.” The work is estimated between $120,000 and $180,000. Of course, there’s a guarantee on the work and naturally, given the tech adjacent nature of this work, cryptocurrency is accepted. By the end of the next week, one lucky buyer will be going home with a patchy brown, five by eight foot canvas. But what will they have really bought?

If you’re unfamiliar with the artist’s CV, I’ll run you through the highlights. Created in 2019, Ai-Da is a vaguely human-shaped robot with a strangely waxy face, full makeup, two distinct metal arms, no legs, a wig with all the realism of Andy Warhol’s and a terrible sense of fashion. Its hardware was built by students at Leeds University and Cornwall-based animatronic fabricators Engineering Arts and its software was designed by PhD students from the Machine Learning Department at the University of Oxford. Ai-Da first exhibited at ‘Unsecured Futures’ at the Barn Gallery, St John’s College, University of Oxford and has since shown at Annka Kultys Gallery, The Design Museum, König Galerie, the Barbican, the Tate, the V&A, the United Nations and the Ashmolean Museum.

They’ll be buying a representation. A mixed media painting based on a famous photograph of Alan Turing, the ‘father of modern computing’ painted in Ai-Da’s signature ‘fragmented’ style: curving brown strokes and an attempt at Pointillism from two perspectives. According to the engineer, the technique is based on early 20th-century masters including Pablo Picasso and Max Beckmann. How exactly it draws on these artists and how the software produces these images isn’t clear. The mind of Ai-Da is closed source and unknowable.

One thing’s certain, it isn’t the expression of a personal vision. Ai-Da is after all, a machine. The press releases say that thanks to the AI algorithms, no two pictures are the same. But they are still the result of two digital camera ‘eyes’ processed by code, transmitted by wires and drawn by mechanical hands. All of which were developed by humans. And those results are… shaky at best. Look at Ai-Da’s other works. She has slow and stilted movements and an unsteady hand. The scratchy pen drawings look like the output of a dot matrix printer or, if you’re being generous, like Frank Auerbach on a bad day.  In terms of output, Ai-Da is hardly a virtuoso. Some of her small paintings, the kind of works that other artists might call sketches or studies, take more than five hours to complete. Not as slow as Lucien Freud, but still, Ai-Da is not the kind of artist to dash off a work on a restaurant napkin. Even if you’re interested in her work, it can’t be a pleasant experience sitting for her, and it’s not like you could drink and smoke together, or spark up an affair.

That said, Ai-Da is rooted in human interaction. Buried in the press releases and some press coverage are references to Ai-Da ‘collaborating’ with humans on artworks. The process goes like this: Ai-Da makes some sketches, then an artist (previously, Suzie Emery) applies the paint, and Ai-Da finishes off the work with a few final flourishes. This kind of production line painting is common throughout art history, but it feels a little odd when it’s a machine taking the credit. In the case of A.I. God. Portrait of Alan Turing (2024) Sotheby’s lists only Ai-Da and Aidan Meller as the creators of the work, presumably they are buying a fully artificially produced artwork? All this is supposedly to “encourage discussion about the relationships and dynamics between machines and humans.” You’d think that with 67% of the world already logged on to the World Wide Web, that discussion has already taken place.

Are they in fact buying into conceptual art? Contemporary art isn’t about the object. Like most contemporary artists, Ai-Da is better at creating hype than canvases. Ai-Da does so by addressing ‘issues’. She says the right stuff – nothing special, just the right words (artificial intelligence, algorithms, creativity) to catch the attention of people who should know better. What Ai-Da thinks of these issues (I’m saying it again, she is a legless robot, she can’t think about these issues) is never answered. Yet, Ai-Da has managed to grab attention like few other human artists. When she addressed the House of Lords, she stood (in effect) in the second chamber of parliament and responded to a question from Baroness Featherstone on the impact of technology on the creative industries. How many human artists have done the same? In her answer, she didn’t have much new to tell us about human-robot interactions. Except that prominent people, lawmakers among them, can be wowed by a shiny toy.

Maybe they’re buying something else… a sexist fantasy. There’s a lot to unpack here. Do they want to have sex with the robot? As many others have noted, it’s interesting that Ai-Da takes the form of a young female artist. She attracts some (Waldemar Januszczak) and intrigues others (Imogen West-Knights), while many more (Naomi Rea, Dan Fox) are disgusted by what they see as objectification. Supposedly, the appearance was to convince young women to study tech. The fact she looks like a Realdoll wrapped around a Boston Dynamics robot is just a coincidence. Would they feel the same about Ai-Da and her art if she was an old, depressed and slightly bloated male robot?

Have they bought into the future? Every buyer buys into an artist’s future, or the future of art. It’s a tacit endorsement of what’s going to remain interesting and in-demand. Is Ai-Da the future? I’m pretty certain when I say no. Despite the hype, Ai-Da doesn’t represent a great technological leap forward. Her paintings don’t represent anything new in art history. They’ve bought into the world as it is. The world of low expectations and failed promises.

Tech was supposed to save us: how many years have we been told that ‘self-driving cars’, ‘manned mission to mars’, ‘carbon capture’ and now ‘artificial general intelligence’ are ‘just around the corner’. When Elon Musk first unveiled his first Tesla ‘Optimus’ robot in 2021, it was a man in a bodysuit. Earlier this month, when he had the ‘Gen 2.0’ versions paraded in front of the press and tending bars, it was soon revealed that they were being remotely controlled by humans.

Ai-Da isn’t quite as big a disappointment. She is, by many measures, an artist. Recall your art history education: Ai-Da matches the representational theory of art, the expressive theory of art, and the institutional theory of art: she reproduces images, she reflects the world around her, and she’s accepted by those inside the Art World. She wouldn’t even be the first artist to be backed by a team of assistants, and plenty of other artists have made a career on being hot and reliably giving quotes to camera. There are many awful artists in the world. Of them all, she might be one of the best.