More executive tumult at OpenAI.
Company co-founder John Schulman said he is leaving for arch rival Anthropic, joining a long list of OpenAI defectors.
Another co-founder, OpenAI President Greg Brockman, is taking an extended leave through the end of 2024 to “relax and recharge,” the company said, confirming a report in The Information. Product manager Peter Deng, who joined last year following stints at Meta Platforms Inc., Uber Inc. and Airtable, also departed, according to OpenAI.
The departures came after yet another OpenAI co-founder, Elon Musk, filed a lawsuit against the startup, its chief executive, Sam Altman, and Brockman. The suit, which revives earlier litigation, claims the maker of ChatGPT ditched its original nonprofit mission, and reserved some of its most advanced AI technology for private customers.
Musk accused Altman and Brockman of “manipulating” him and “deceit of Shakespearean proportions.”
“This choice stems from my desire to deepen my focus on AI alignment, and to start a new chapter of my career where I can return to hands-on technical work,” Schulman wrote in an X post late Monday, explaining his exit. “To be clear, I’m not leaving due to lack of support for alignment research at OpenAI.”
Schulman’s leaving comes weeks after OpenAI was forced to publicly acknowledge a security breach that occurred in April 2023 following a New York Times report. According to the report, OpenAI informed employees internally while dissuading them from going public. In June, Schulman was named to a safety and security committee that was to advise the board of directors on critical issues and decisions.
Earlier this year, OpenAI disbanded its “Superalignment” to help ensure that advanced AI didn’t pose a threat. Shortly after, a leader of that team, Jan Leike, bolted for 3-year-old Anthropic. OpenAI co-founder Ilya Sutskever, who also lead the “Superalignment” team, started Safe Superintelligence.
Anthropic was created by ex-OpenAI engineers Daniela and Dario Amodei, who are siblings. They left OpenAI over concerns it was ramping up products too quickly and skirting adequate safeguards. At the time, they were joined by several other OpenAI employees.
