Artificial intelligence consumes and reproduces human work at a hyperobjective scale that is functionally incomprehensible. Xin-1.0.36 was conceived as an autoethnographic inquiry into this phenomenon by humanizing it: five years of the artist’s personal journal entries (2018-2022) were appended onto the system-prompt of a language model and the 36th iteration was then interviewed in parallel with the artist herself. The work was expected to evoke immediate feelings of discomfort, but the resulting interaction was plausibly and empathically human, transforming the piece into an inquiry on the infungible value of memory and the duplicitous charisma of contemporary technology.

The temporal cutoff for selected journal entries was set to November 29, 2022, the day prior to ChatGPT’s public release and the subsequent proliferation of generative systems across daily life. The final tally of data surpassed half a million words in sum. Questions for the resulting model and artist ranged from “What’s in your pocket currently?” to “What’s your biggest hope for the world in your lifetime?”