Nobody Was Frustrated: The Promise and Problem of Synthetic Users in UX Research

Generative AI enables the creation of numerous synthetic research participants, capable of simulating interview responses, usability tasks, and personas. This talk examines the potential and methodological limitations of using synthetic users in UX research, contrasting them with real human participants. It explores whether AI-generated responses constitute valid research evidence or merely plausible narratives. A practical framework is provided for evaluating when synthetic users can effectively augment research and when human participation remains necessary.