- John Schulman, a founding member of OpenAI, is leaving for Anthropic PBC.
- OpenAI’s founding team now includes CEO Sam Altman, President Greg Brockman, and researcher Wojciech Zaremba.
- Brockman announces a four-month sabbatical following Schulman’s departure.
- Schulman’s work at OpenAI focused on reinforcement learning and AI safety.
- Anthropic, a competitor, has raised over $7 billion and boasts advanced AI models, including Claude 3.5 Sonnet.
- Jan Leike, another former OpenAI researcher, previously joined Anthropic.
- Peter Deng, a recent product manager at OpenAI, has also left the company.
Main AI News:
John Schulman, a pivotal figure in OpenAI’s founding team and a key contributor to its product development, has decided to leave the company and join Anthropic PBC, a direct competitor. His departure, announced today on X, leaves OpenAI with three original founders: CEO Sam Altman, President Greg Brockman, and researcher Wojciech Zaremba. Brockman, in response to Schulman’s exit, has revealed his plan to take a four-month sabbatical.
Schulman, who came to OpenAI in 2015 after earning his PhD in computer science from UC Berkeley, has been influential in developing AI technologies. His work on reinforcement learning has played a significant role in advancing OpenAI’s models. Schulman co-led the post-training team at the company, which is essential for refining AI models and implementing safety measures.
As the head of alignment science, Schulman ensured AI models were equipped to avoid generating inaccurate or harmful outputs. His move to Anthropic, which has raised over $7 billion in funding and is known for its Claude series of language models, represents a notable shift. Anthropic’s latest model, Claude 3.5 Sonnet, is said to outperform OpenAI’s GPT-4o across various benchmarks.
Anthropic’s recruitment decision has been strategic, with the company successfully attracting Jan Leike, another former OpenAI researcher specializing in AI alignment. Leike’s previous comments on the challenges at OpenAI led to the formation of an AI safety and security committee that included Schulman.
Brockman’s sabbatical, announced shortly after Schulman’s departure, is his first significant break since co-founding OpenAI nine years ago. Despite these leadership transitions, developing a safe AGI remains a priority.
Additionally, it was reported that Peter Deng, who joined OpenAI in 2023 as a product manager, has also left the company. Deng had previously worked at Meta Platforms and other major tech firms.
Conclusion:
John Schulman’s departure from OpenAI to Anthropic is a significant event that signals a shift in the competitive landscape of the AI sector. As OpenAI’s founding team undergoes changes, including Greg Brockman’s sabbatical, the company faces the challenge of maintaining its innovation momentum. Schulman’s move underscores the ongoing talent migration between leading AI firms, with Anthropic gaining from increased expertise in AI alignment. This transition could bolster Anthropic’s competitive advantage and influence OpenAI’s strategic positioning, underscoring the dynamic nature of the AI market and the importance of retaining top talent.
