I felt a vague sense of disquietude when I received word from Christie’s that it would be hosting the first-ever artificial intelligence-dedicated sale at a major house. Just a little moment of unease. Frankly, it’s hard to shush those quiet biases that make a person want to categorize all art into “real art” and “other,” as if I have somehow earned the right to make that distinction. But unpacking my kneejerk reaction to the Augmented Intelligence auction, which will be open for bidding from February 20 through March 5, didn’t take long. I am a person who makes things, and the adoption of artificial intelligence technologies in creative spheres feels like an existential threat when you’re someone who has roped material survival to artistic impulses.
I’m also a realist. Artificial intelligence can write and draw and, tethered to a robotic arm, it can paint, and it can make data so beautiful we put it in museums, but there’s always a human being behind the curtain pulling the strings, whether by writing code or dreaming up ideas or otherwise telling A.I. what to do. (For the record, it can also chase glitches, sus out breast cancers better than human doctors and, just maybe, break your porn habit.) I also futz around with artificial intelligence on the regular because I’m not as afraid of A.I. stealing my job as I am of someone who’s really good at using A.I. stealing it.
When it comes to art, the question we like to think we’re grappling with—beyond how we’re going to pay the bills when ‘content’ supersedes art across all genres of creative output—is just how much human intervention is necessary to make A.I. generated art palatable to the masses. But some small, possibly also biased studies seem to find that most people actually like A.I. art as long as they think it was made by a human being. Probably because most average Joes want art that looks nice, and today’s artificial intelligence tools are more than capable of generating nice-looking pictures, no matter who is at the helm. That’s part of the problem, of course. The folks shouting from the virtual rooftops that A.I. art is hot garbage might reference copyright infringement (a very real concern) and artist livelihoods (same), but what it often comes down to is that nagging feeling that using artificial intelligence to make art is automating creativity. Replacing effort. Cheating. Too easy.
I should point out that the lots in the Augmented Intelligence sale are described as being created with and not by A.I., which I feel is an important distinction even as I’m not sure where the line between with and by actually lies because I’m a sometimes Luddite and after a while, I lose the thread. But among the artists with works in the Christie’s A.I. sale are recent TIME100 A.I. Impact Award recipient Refik Anadol, early A.I. art pioneer Harold Cohen and Alexander Reben, OpenAI’s first artist in residence—none of whom seem like corner-cutters. “The longer it takes to create something, the better it is, whether it be wine, whisky or a work of art,” asserted author, publisher, former pro skateboarder Scott Bourne in a 2024 Observer article that explored whether A.I. is a creative supertool or art imitator. I don’t think he’s necessarily wrong about the impact of time on quality, but I also don’t think he’s right to conflate A.I. art with speed. Cohen, for his part, spent decades working on AARON, widely considered to be the earliest A.I. art software. Put another way, art with extra steps.
SEE ALSO: At Regen Projects, Doug Aitken Reimagines Nature Through the Lens of the Anthropocene
Josh Tyson, who interviewed Bourne, wrote that “generative A.I. has kicked over the barrier of technical training as a means to create art.” Here I also have to disagree. Learning to wield a paintbrush is one type of technical training; learning to build datasets is another. I look at what Alexander Reben is doing, and I certainly wouldn’t call it easy or a shortcut to creativity. And what is real art, anyway? “I get excited when people think I’m not making ‘real art’ because it means I’m doing something different and writing my own story,” Sougwen Chung told Observer last year.
Meanwhile, an open letter calling for Christie’s to take down the auction has amassed more than 3,500 signatures. “Many of the artworks you plan to auction were created using A.I. models that are known to be trained on copyrighted work without a license,” said the letter’s anonymous authors. “These models, and the companies behind them, exploit human artists, using their work without permission or payment to build commercial A.I. products that compete with them.”
The letter doesn’t point to specific lots created with suspect tools, and Anadol, in a tweet, countered that the “majority of the artists in the project [are] specifically pushing and using their own datasets [and] their own models!” Some countered that we shouldn’t dismiss the widespread anxiety about artificial intelligence among working artists and creatives who are worried about their livelihoods drying up. Still others pointed out that A.I. is simply a data-powered method of achieving a task. If the sale had been called the Neural Net auction, would people be reacting as strongly?
The easy-way-out argument is nothing new, by the way. Charles Baudelaire, in The Mirror of Art, wrote that “the photographic industry was the refuge of every would-be painter, every painter too ill-endowed or too lazy to complete his studies,” but more than 160 years later, there are still thousands of phenomenal painters out there putting brush to canvas. The Christie’s sale will, according to Nicole Sales Giles, the auction house’s VP and director of Digital Art, show “collectors human agency in the age of A.I. in fine art.” But you don’t have to be a collector to see for yourself and come to your own conclusions, which might differ from mine. The various sculptures, paintings, prints, works on paper, digitally native works, screens and interactive works will be on view at Christie’s Rockefeller Center galleries starting on February 20.
