Co-founder John Schulman will work on AI alignment at Anthropic
Less than a day after one OpenAI co-founder filed a lawsuit against the startup and key executives, its president announced a months-long sabbatical and co-founder John Schulman revealed he would be joining rival artificial intelligence firm Anthropic.
In addition to helping launch the Microsoft $MSFT-backed startup, Shulman led OpenAI’s post-training team, which fine-tuned the models used to power AI chatbot ChatGPT. In June, Shulman was appointed to a safety and security committee designed to provide advice to the board of directors on critical safety and security decisions.
“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 a post on X $TWTR, formerly Twitter, Monday evening. “To be clear, I’m not leaving due to lack of support for alignment research at OpenAI,” he added.
Earlier this year, OpenAI disbanded its “Superalignment” team, which was focused on controlling AI systems and ensuring that advanced AI wouldn’t pose a danger to humanity. One of the team’s leaders, Jan Leike, joined Anthropic shortly after his group was axed. Ilya Sutskever, another Superalignment leader and an OpenAI co-founder, started his own firm, Safe Superintelligence. Another co-founder, Andrej Karpathy, has also left the company.
Anthropic was launched in 2021 by ex-OpenAI engineers and siblings Daniela and Dario Amodei, who left that startup over concerns that it was scaling up its tech too quickly — and without adequate safeguards. They were joined by at least nine other OpenAI employees.
Schulman’s announcement was met with appreciation for his time at OpenAI; CEO Sam Altman wrote on X that Schulman’s perspective on the future of AI helped inform the company’s strategy. “Very excited to be working together again!” Leike wrote in a reply to Schulman.
OpenAI President Greg Brockman on Monday announced he would be taking a sabbatical from the company through the end of 2024, writing on X that it would be his “first time to relax since co-founding OpenAI 9 years ago.” Both executives’ moves — as well as the departure of product leader Peter Deng — were first reported by The Information.
Join 500,000+ readers who start their day with Quartz.
By subscribing, you agree to our Terms of Service and Privacy Policy.
