Paul is a technical advisor to ARC Evals as well as a board member. He runs the Alignment Research Center, a non-profit research organization whose mission is to align future machine learning systems with human interests. ARC Evals started as a project of Alignment Research Center, and is now spinning out from ARC.
Paul previously ran the language model alignment team at OpenAI; his work included leading the research on the Deep reinforcement learning from human preferences paper. He has a PhD in Computer Science from the theory group at UC Berkeley.