Envisioning and Understanding Orientations to Introspective AI: Exploring a Design Space with Meta.Aware

Introspection is the practice of looking inward for ongoing selfexamination. It involves considering one's past experiences and asking questions about the present and future. Our work investigates how AI could open new possibilities for supporting introspective experiences. Adopting a design fiction approach, we created a fictional company called Meta.Aware to contextualize 4 different Introspective AI product concepts in the form of video sketches. We used the Meta.Aware platform to conduct interviews with 17 participants, using the 4 concept videos as prompts for discussion. Participants had a range of reactions related to perceived benefits and tensions in this emerging design space. We interpret these results to outline future design directions for mobilizing AI as a resource to support introspective experiences over time, as well as to reflect on issues and dilemmas bound to this emerging design space.

• Human-centered computing → Interaction design process and methods.