The good news is that I’m a journalist, so I asked some artists, researchers and art critics what they thought of the aesthetics of AI art. First I called Amelia Winger-Bearskin, an artist and professor at the University of Florida. Winger-Bearskin has been catalog several visual trends she has noticed in recent AI art. She mentions a trend Nightmare Corp.— often illustrated by images conjured up by Google’s Deep Dream, an older generator released in 2015. It specializes in swirling, psychedelic imagery, such as memories of a particularly poignant acid trip. “Prog-rock influences, for sure,” she says. Another category Winger-Bearskin explores, which she calls Dada 3D, is much like the crazy scenes I conjure up when I play with these generators. She describes it as “something like a surreal parlor game.”
In addition to taxonomizing trends, Winger-Bearskin has noticed broader stylistic tics in these generators. She sees Western animation and Disney-style anime as obvious influences, as well as a tendency to treat whiteness as the default race – a result, she suspects, of training these generators on datasets heavy for Western animation at Disney style, anime , and images of white people.
Lev Manovich also pays close attention. The cultural theorist and professor at the City University of New York has been lurking on Midjourney’s Discord server since last year to analyze how people use the generator. After Midjourney released an update last fall, he saw some changes to what people were urging the generator to make. For example, after it got better at portraying people realistically, requests for portraits of both men and women went up.
Digital artist Sam King first began to closely monitor the AI art scene in 2021. Excited by what they saw, they began parts their favorite photos on social media, building a following as a curator just as technology was taking off. They describe the earlier wave of generators as favoring “trippy, abstract stuff.” (These generators are known as Generative Hostile Networks or GAN. I have a Couple of people call this look, rather uncreatively, GANism.)
King considers the latest generation of generators, called diffusion models, to be stylistically distinctive. Just as oil paintings and watercolors produce recognizably different effects, GAN generators and diffusion generators produce recognisably different images. If you want a more realistic depiction of, say, Tony Soprano drinking a cappuccino with Shrek, the diffusion models are more likely to yield convincing results. “In theory, you can create all sorts of different aesthetics with these machines,” they say. However, more realistic does not necessarily mean more stylistically varied. Like Winger-Bearskin, King often sees Disney and anime influences popping up, as does comic book art.
“The rhetoric of these companies is that you can make anything you can imagine. It’s about this open border. But of course popular culture follows certain stereotypes and tropes,” says Manovich. He keeps seeing variations on different themes: “Fantasy, fairy tale, comic book, video game.”