Queering AI: Undoing the Self in the Algorithmic Borderlands

This paper challenges fixed orientations towards the self in human-AI entanglements. It offers queering as a strategy to subvert the individuation and fixing of identities within algorithmic systems and the loss of futurity that it brings about. By exploring queerness, the paper examines how one's sense of self and futurity are interpellated within the algorithmic borderlands of human-AI entanglements. The study discusses an embodied experiment called "Undoing Gracia," a Digital Twin simulation where the first author Grace and their AI twins (Lex and Tortugi) interact within the fictional world of Gracia. The experiment probes into Grace's multifaceted subjectivities by conceiving themselves as interdependent entities evolving through their interactions within Gracia. The paper outlines the process of creating and implementing the simulation and examines how the agents co-perform and become-with alongside Gracia's making. The findings illuminate queer gestures for navigating human-AI entanglements in HCI research and practice, highlighting the importance of fluid identities in shaping human-AI relations.