The 'Godfather of AI' backs a new watchdog plan to track OpenAI and Anthropic's AI risks from the inside (www.businessinsider.com)

The AI Evaluator Forum has released a letter outlining essential conditions for AI labs like OpenAI and Anthropic to enhance collaboration with third-party evaluators who assess technology risks. This initiative is backed by prominent figures in the AI field, including Geoffrey Hinton, known as the "Godfather of AI," and Stuart Russell, following commitments from both companies to welcome outside evaluators to gain deeper insights into their AI safety practices. Significantly, the letter emphasizes that evaluators must maintain scientific objectivity, transparency, and independence, with the ability to communicate freely with boards and oversight bodies. The proposal includes granting evaluators access to the same systems and data as internal assessors and encourages diverse expertise to address various risks, such as those related to biological sciences and loss of control. This collaborative approach aims to foster accountability and trust in AI development amidst rising concerns about the technologies' implications.