John Schulman, a co-founder of OpenAI, has departed from the organization to join Anthropic, a competing artificial intelligence startup
In addition, the company has announced that Greg Brockman, the president and co-founder of OpenAI, is taking an extended leave of absence until the end of the year. Brockman has been with the company for nine years and uses this time to “relax and recharge.”
Peter Deng, a product manager who joined OpenAI last year after overseeing products at Meta, Uber, and Airtable, also exited the company some time ago, the company confirmed.
The Information had previously reported the news of Brockman and Deng’s departures.
A spokesperson for OpenAI issued the following statement regarding Schulman: “We appreciate John’s contributions as a founding team member and his unwavering dedication to the advancement of alignment research.”
His dedication and diligence have laid a solid foundation to inspire and incentivize future advancements in OpenAI and the broader field.
Today, Schulman announced his decision on X, citing a desire to intensify his interest in AI alignment, the field of science that ensures AI behaves as intended, and to participate in more hands-on technical work.
Schulman stated, “I have resolved to pursue this objective at Anthropic, where I am confident that I can acquire novel perspectives and conduct research in collaboration with individuals who are profoundly invested in the subjects that interest me.” “I am certain that OpenAI and the teams I was a part of will continue to flourish in my absence.”
Schulman’s involvement with OpenAI commenced immediately after he completed his Ph.D. in electrical engineering and computer sciences at UC Berkeley.
He was instrumental in developing the AI-powered chatbot platform ChatGPT by overseeing OpenAI’s reinforcement training organization, which refines generative AI models to adhere to human instructions.
Schulman assumed the role of the “post-training” team, also known as the alignment science endeavors of OpenAI, following the departure of AI safety researcher Jan Leike, currently employed at Anthropic.
He was also a member of the safety committee that OpenAI recently established; it is uncertain who may assume Schulman’s position.
Schulman stated that he was not departing OpenAI due to a lack of support, despite the controversies surrounding the company, particularly its approach to and treatment of AI safety research.
“Company leaders have highly prioritized the investment in [alignment research],” Schulman stated. “My decision is personal, contingent upon my intentions for the subsequent career phase.”
After Schulman’s departure, the number of founders at OpenAI has been reduced to three: Wojciech Zaremba, the chief of language and code generation, Brockman, and OpenAI CEO Sam Altman.
In a post on X, Altman expressed his gratitude to Schulman for his contributions to OpenAI.”You are a brilliant researcher, a deep thinker about product and society, and, most importantly, a great friend to all of us.” We will miss you terribly and strive to make you proud of this location.
