Talking to Oneself With and Through AI

Talking to Oneself With and Through AI In this talk, I explore the possibilities and limitations of using AI chatbots for talking to and with oneself. Marya Schechtman's (2026) recent work takes the example of Michelle Huang’s artistic experiment of feeding in her old diaries to ChatGPT in order to talk to her younger self, and considers what kind of self-knowledge might emerge from such an interaction. Taking Schechtman's work as a launching point, I consider more prosaic ways in which engaging with a chatbot might give rise to distinctive forms of self-engagement and self-dialogue. In doing so, I shift the attention from self-knowledge to the ambiguous phenomenology of the experience itself. In particular, I bring a perhaps unexpected guest to the table. Drawing on Hannah Arendt’s work on the importance of solitude (understood as being with oneself as a ‘two-in-one’) for critical thinking and reflection, I consider what kind of self-dialogue interacting with AI might facilitate and whether it can scaffold the self-dialogue Arendt deems central to this essential practice.