The nonprofit OpenAI Foundation has added a figure well known beyond the company to its board of directors: Paul Christiano, a researcher in AI alignment and safety. TechCrunch reports this, according to infohub.kz.
In addition to serving on the foundation's board, Christiano has taken a seat on OpenAI's safety and security committee. The decision comes amid heightened public and regulatory scrutiny of oversight and safety for advanced language models.
Paul Christiano previously led the alignment team at OpenAI. He is credited with a key contribution to the development of reinforcement learning from human feedback (RLHF), a method now used in building popular neural networks.
He later founded the research organization Alignment Research Center (ARC) and has served as an adviser to the U.S. Center for AI Safety and Innovation (CAISI).
Christiano gained prominence for warning about the potential risks of uncontrolled AI and the possibility of an existential threat to humanity from the development of superhuman intelligence.
Observers note that bringing in a respected safety specialist should strengthen OpenAI's corporate governance and restore confidence in its policies for preventing risks associated with the development of strong AI.