Large Language Models are pervasive in modern life. We use them to draft e-mails, write essays, search for information, and sometimes seek advice. In short, an increasing number of our thoughts and decisions are informed by the outputs of these AIs. It is therefore important for users, designers, and the general public to be aware of the values embedded in this technology.

This site presents the research results of the Philosophical Hub for AI Research on Society (PHAROS) at the University of North Carolina, Wilmington. We seek to shed light on the values expressed by LLMs by employing methods used in philosophy and psychology.

This research was generously funded by the Cosmos Institute.