Yesterday
@aza and I were honored to interview William Saunders, who was part of the recent group of OpenAI whistleblowers who came forward.
podcasts.apple.com/au/podcas…
William worked on safety issues at OpenAI but resigned because he saw the company consistently prioritize growth and market dominance over the safe roll out of AI.
He co-wrote the
righttowarn.ai/ letter and personally risks losing millions to warn us. The letter was also endorsed by Yoshua Bengio, Stuart Russell and Geoff Hinton. With the lack of meaningful regulation of AI currently, whistleblowers are one of the few ways society can have early warning to systems that pose risk. We need independent oversight of AI companies, and should explore ways of giving company insiders the right to warn us of the risks that only they can see.