OpenAI CEO Sam Altman told a conference Tuesday that AI companies should expect accidents as increasingly capable models are deployed, and urged the industry to build a culture of openly reporting and learning from them, adding that he is confident the industry can police AI safety.
“Regrettably, accidents with any new technology are unavoidable,” Altman said during an interview with Salesforce CEO Marc Benioff at Dreamforce.
Altman compared the approach he wants the AI industry to take to aviation, where regulators and companies investigate accidents and use those findings to improve safety of the broader system.
