David Robinson, who said he spent three and a half years at OpenAI, resigned and wrote that the company was not careful enough before releasing more capable systems.
A former OpenAI safety employee, David Robinson, resigned and criticized the company's approach to AI safety in an essay published by The Atlantic on Saturday. Reuters reported the essay under the headline that the time for trial and error is over.
Robinson wrote that AI companies, including OpenAI, were not being nearly careful enough and should put more weight on safety expertise and research before building more capable systems. He argued that advanced systems need safeguards closer to those used in nuclear power and aviation.
He said he spent three and a half years at OpenAI, helped draft the company's preparedness framework, and oversaw safety reports for 12 frontier-model launches. He wrote that as the company sprints from one launch to the next, it is failing to achieve the level of care he believes is needed.
Robinson criticized what OpenAI calls iterative deployment, the practice of releasing systems and tightening safeguards after problems appear. He also said capabilities were advancing faster than researchers' understanding of alignment.
An OpenAI spokesperson said the company was making sure models do not become more capable than it can safely manage and secure, and that it pauses training or holds back models when it needs to slow down.
People Sentiments Negative
- Robinson wrote that the time for trial and error is over
- An OpenAI spokesperson said the company pauses training or holds back models when it needs to slow down