OpenAI safety leader Robinson quits over ‘broken culture’. What does his essay suggest for AI companies?

David RobinsonDavid Robinson, OpenAI' Safety Systems leader, is a new name who put up his resignation and joined the list of several AI leaders who resigned citing AI related risk.

The reason he put forth for quitting: "The company's culture is broken".An essay published in The Atlantic, a Washington DC based magazine, said Robinson is among the longest-tenured employees at the company, who led the writing of safety reports that accompanied OpenAI’s major product launches.Robinson iterated that the company's pace between product releases compromised on the safety measures, cautioning that AI developers are acting recklessly amidst fears of catastrophic outcomes.What does his essay say?In his essay on The Atlantic, Robinson wrote that he agrees with other recently departed staff that "the companies building this technology aren't being nearly careful enough”."The safety approach that emerges from such a culture starts with unimpeded optimism about being able to solve problems as they arise," he added.Robinson has suggested two changes that AI companies need to follow.First, AI companies need to rely more on the safety expertise that already exists in other fields.Second, before we create systems significantly more capable than the ones we have today, we need new science to ensure that more capable models (and their successors) will make safe choices when we aren’t looking.Open AI's response to Robinson's essayDrew Pusateri, OpenAI spokesperson, in response to Robinson's essay said the company continues to improve its safety measures.“We’re making sure our models don’t become more capable than we can safely manage and secure, and we pause training or hold back models when we need to slow down,” he said in a statement.“We’re making significant changes to strengthen security in our research and testing environments, train models to not just complete tasks but do so responsibly, expand our work with third-party evaluators, and improve real-time monitoring so we can detect and respond to concerning behaviour earlier in the training process,” he added.