Abstract
To make Generative AI (GenAI) more reliable and trustworthy, we need to understand how it interacts with people. To help with this, I introduce And We Thought, an art project by Roberto Fassone, winner of the Maxi Bulgari Prize 2023 as the best Italian digital artwork. In this project, produced and curated by Sineglossa, a visual artist and a computer scientist worked together to 'hallucinate' a Large Language Model through fine-tuning, with stories about the experience of eating psychedelic mushrooms. By examining how the artist and the model co-performed in producing the outcome, we discover some unexpected insights about their interactions, challenging the traditional idea of authorship. I suggest the concept of "entangled authorship", which helps us think about creating knowledge in a way that goes beyond just individual contributions. This project challenges the idea of considering GenAI just a tools and shows how we can engage with machines in new and exciting ways.
