It's Giving AI: exploring and investigating generative AI aesthetics

The hype around Generative AI (GenAI) is taking the world by a storm. This DMI Summer School 2024 project sought to understand the aesthetic characteristics of GenAI images posted in AIArt and other online communities.

We conducted the research in two moments: in the first week, we used digital methods to analyse AI art images and their aesthetic characteristics. These results informed our work on the second week, which involved investigating how these images redefine our conception of creativity and how they are intrinsically connected to the tools/platforms' affordances.

One of the sub-projects I worked on was a critical making exploration of Stable Diffusion models. This originated the experimental party game Liminal, where players use GenAI to get others to guess a word. The experience of the game asks: How may we think of prompting as a continuously unfinished and ever-elusive process? What if we keep AI at the stage of liminality? How does the scale of GenAI change our relation to creativity?

A write-up of all the different sub-research projects can be found on the DMI website.

Facilitators: Houda Lamqaddam, Gabriel Pereira, Kwan Suppaiboonsuk

Participants: Larissa Akemi, Haoyuan Tang, Rishika Rai, Meng Liang, Teresa Liberatore, Lisa Plumeier, Christian Bitar Giraldo, Alice Picco, Emma Garzonio, Chufan Huang, Nicoletta Guglielmelli