By Natalia Gonzalez Blanco Serrano ’24
Famous surrealists of the 20th century often engaged in an activity dubbed “exquisite corpse;” dividing a paper into thirds, different artists would draw whatever they liked without seeing each other’s contributions. The result was often a wacky, indecipherable mix of styles, situations, and body parts that pushed and stretched surrealist artists beyond the confines of their own imagination. This was the inspiration behind DALL-E, a new artificial intelligence art-generating tool and subject of a spring quarter display at University Libraries.
“Instead of looking at (AI) as a threat, I was trying to look at it as an opportunity for collaboration, like an expansive arts practice,” said Kat Caribeaux, art history PhD candidate and innovator in residence at the Emerging Technologies Lab.
DALL-E generates images based on a text prompt input by users. While Caribeaux understands many of the current drawbacks and hesitations around such AI generated images, she believes we should be using these tools to expand our conceptions of art and the tools we can use to make it. On her first day as a TA in art history, she asked her students to use DALL-E to generate a piece of modern art and improvise an artist statement.
“I wanted to do something that was a little silly to kind of break down like the intimidation factor of art history,” she said. “But a second part was to get students paying attention to visual trends within modern art.”
In “Artificial Intelligentsia,” which opened at University Library March 29, Caribeaux and the Emerging Technologies Lab collaborated with five Helicon literary magazine writers and DALL-E to continue innovating in art. Twenty-five images 
generated by this process are on display in the exhibit.
“We had the writers in for a workshop, where they came in with some prose, and we helped them translate it into prompts that DALL-E would understand,” Caribeaux explained. “DALL-E then would produce about four images, and then they’d select one of those images to inspire the next piece of prose.” These images range from sepia-toned photographs to abstract works of art.
The Emerging Technologies Lab’s collaboration with Helicon is just the beginning of AI-art collaboration, but it doesn’t come without a sense of responsibility. “Sometimes when folks think about emerging technologies it can be really technophilic, but that’s not our approach at all,” said Caribeaux. “[At the Emerging Technologies Lab] we feel like we really do have an ethical and social responsibility that our applications of these emerging technologies are doing broad good to the people they’re being brought to.”
Natalia Gonzalez Blanco Serrano is a Medill School of Journalism junior