Appearances and Expectations in Human–AI Interaction

• Department of Political and Social Sciences

PhD thesis defence by Jonathan Winter.

From chatbots to humanoid robots, AI agents are entering social roles once occupied only by humans. Increasingly designed with human-like appearances, names, and voices, these agents are reshaping how people navigate social life. This dissertation examines how demographic appearances and social expectations shape behavior across human and human-AI interaction. Four empirical studies use controlled experiments and large-scale observational data to trace how people select and collaborate with human and AI partners. The studies examine whether demographic preferences extend to AI agents, whether algorithmic transparency can reduce appearance-based bias in real-world selection, and how teaming with AI affects individual motivation and performance. Together, the findings shed light on how social behavior adapts to a world in which humans are no longer the only social agents.

Jonathan Winter studies how people perceive and respond to AI agents designed to occupy social roles. His research examines topics such as partner selection, team dynamics, social bias, and AI companionship, with a broader interest in how artificial agents reshape human judgment and interaction. Beyond academia, Jonathan has held applied research roles in the technology sector and at policy institutions, where he studied the social and economic implications of AI.