make sense of things [A]

performance, 2024

This is a sequence of recurring circular conversations with the voice mode of Open AI's Chat GPT.

With a database not updated since 2022, it is not aware of the recent escalation of the long-standing violence we are witnessing. Though the AI claims to not be able to form its own opinions, it routinely illustrates its own biases in our conversations: maintaining a passive stance of complexity when speaking of historical contexts to this present moment + demonstrating avoidance when asked whether cutting off water for a community is a criminal act. With each conversation, there has been an increase in the AI's evasion to assign a clear perpetrator. I discovered in one of our chat logs that some of my questions had been flagged as going against Open AI's content policy.

As the floor gets more wet things begin to slip. This piece is interested in how truth is formed and assigned, what is programmed, and what deprogramming is possible.