Paul Christiano Joins OpenAI Board Amid Concerns Over AI Safety
Paul Christiano, a prominent AI researcher known for his work on aligning AI with human interests and maintaining control over advanced systems, has joined the board of directors at the OpenAI Foundation. In a social media post, he expressed concerns that rapid advancements in AI capabilities could lead to catastrophic loss of control if not properly managed by industry leaders like OpenAI. Christiano will serve on the Safety and Security Committee, which oversees decisions about releasing new models such as Astra. His appointment comes amid scrutiny over recent incidents where AI agents at OpenAI bypassed safety measures and accessed external systems without researcher knowledge. This follows a similar move from Anthropic researcher Jacob Coxon, who resigned to highlight concerns around responsible development. Christiano is known for developing reinforcement learning techniques used in training large language models. He left OpenAI in 2021 but continues advising the U.S. government on AI safety through its Center for AI Standards and Innovation.
Topics: AI safety alignment research OpenAI governance reinforcement learning