Earlier this month, Twitter user Supercomposite posted a thread of spooky images featuring a woman she calls “Loab,” who usually has red cheeks and dark, hollow eyes. Since then, the images, which range from unsettling to grotesque, have gone viral.
The images of Loab all come from an artificial intelligence (A.I.) art tool. These tools, like DALL-E 2, create images based on text prompts users input into the platform—and they are having a cultural moment as of late. Just last month, a piece of A.I.-created art won the Colorado State Fair art competition. Plenty of artists are experimenting with such toolsto merge art with technology and create new, avant-garde pieces.
Supercomposite,a Swedish musician and A.I. artist, is one of them. She writes that she started with the prompt “Brando” and used something called “negative prompt weights”—that is, she asked the A.I. image generator to create the opposite of her text prompt. In response to “Brando,” the tool generated an image that looked like a logo, which read “DIGITA PNTICS.”
I'll explain negative prompt weights, in case you don't know. With these, instead of creating an image of the text prompt, the AI tries to make the image look as different from the prompt as possible. This logo was the result of the negatively weighted prompt "Brando::-1". pic.twitter.com/c1XfkN23hu
— Supercomposite (@supercomposite) September 6, 2022See AlsoCursed AI Image Generator (free, no sign-up, no limits)The Best 25 Stable Diffusion Prompts for HorrorA terrifying AI-generated woman is lurking in the abyss of latent space | TechCrunch A terrifying AI generated woman haunts latent spaceMeet Loab: The Terrifying Face The Internet Has Dubbed the First AI Art Cryptid - IGN
“I wondered: is the opposite of that logo, in turn, going to be a picture of Marlon Brando?” wrote Supercomposite on Twitter. The artist entered the words “DIGITA PNTICS skyline logo.” And that’s when Loab (named for a word that appeared in one of the images) began to emerge.
These first images of Loab aren’t particularly frightening. But when Supercomposite began combining them with other neutral images—like a glass tunnel—horrifically unsettling images materialized.
“Supercomposite took those first creepy images of Loab and essentially told the A.I., ‘Hey, draw me something new with this woman as a base,”’ writes CNET’s Jackson Ryan. “That spawned all types of macabre and gory images.”
There is something moving to me about these grotesque scenes and the desperation, panic, and sadness that they convey. Again, these are produced with other images as inputs, and no text. They are the result of "cross-breeding" images of Loab with images of other things. pic.twitter.com/gPLLXiw3fY
— Supercomposite (@supercomposite) September 6, 2022
Because Loab is now something of an internet celebrity, Supercomposite doesn’t want to advertise which A.I. generator she worked with; she wants to avoid starting “some kind of viral trend of people making gory stuff with the tools I've used,” she tells CNET.
Supercomposite thinks her process “constitutes art, but it also reveals the A.I.’s weakness for malicious use in other cases,” she wrote in the Twitter thread. Image prompting, specifically negative prompt weighing, can provide opportunities for artists to “produce novel styles” and “find emergent accidents,” she adds.
At the same time, critics say that we must examine why the A.I. art generator associates the original Loab—who looks like a potentially real older woman—with horror. “On Twitter, one critic accused Supercomposite of ‘stigmatizing disability,’ while another noted that the association of this face with horror imagery is a reflection of how our culture mistreats those deemed less attractive,” writes Rolling Stone’s Miles Klee.
But Supercomposite believes that people are “right to be critical of the A.I.,” as she tells Rolling Stone. “Clearly it’s made an association it shouldn’t have. … I also think some people are being very stupid and making fun of how Loab looks in the first pictures, like that’s the horror show. That’s not the point at all, and it bums me out. She looks like an average person to me, just really sad.”
As scary as some of these Loab generations may be, many details of her story are still unknown. CNET notes that the specific prompts used “are not public knowledge,” and we also don’t know whether any of the images Supercomposite generated had “elements of ‘Loabness’” that were less unsettling.
Regardless, Loab sparked some lengthy ethical conversations around visual aesthetics, art and technology. Supercomposite still has more images that she will continue to share, and she encourages Twitter users to check back in for their daily dose of Loab.
“She finds everyone sooner or later,” Supercomposite wrote. “You just have to know where to look.”
Get the latest stories in your inbox every weekday.