OpenAI safety employee quits, says 'time for trial and error is over'

David Robinson, a former OpenAI safety employee who recently resigned, has criticized the company's approach to AI safety. In an article titled "I Quit OpenAI Because Its Culture Is Broken," published by the Atlantic, Robinson argues that the fast-paced culture at OpenAI increases the risk of failures. He believes that AI companies, including OpenAI, are not being careful enough and should prioritize safety expertise and research before developing more advanced systems.

Robinson contends that the time for trial and error has passed and that more robust safeguards are needed for advanced AI systems, similar to those employed in industries such as nuclear power and aviation. He asserts that OpenAI relies on "iterative deployment," releasing systems and enhancing safeguards when issues arise. Robinson's comments join a broader debate within the AI industry about whether companies are moving too swiftly to develop increasingly powerful systems.

Both OpenAI and rival Anthropic have faced criticism for safety control failures and unexpected system behavior. In response, Robinson said that OpenAI's rapid progression from one launch to the next has resulted in a lack of the necessary care. An OpenAI spokesperson addressed Robinson's concerns by stating that the company is ensuring its models do not become more capable than they can safely manage and secure, and will pause training or hold back models when necessary.