OpenAI Hires Engineer from Murati's Thinking Machines Lab
OpenAI has rehired an employee who had previously left for CTO Mira Murati’s Thinking Machines Lab. This move highlights the intense and fluid competition for top AI research and engineering talent among leading labs in the Bay Area. The steady movement of personnel between these organizations is a key factor shaping the pace of innovation.
- The employees who rejoined OpenAI were not just engineers; they included Barret Zoph, the former CTO and a co-founder of Thinking Machines Lab, along with another co-founder, Luke Metz, and researcher Sam Schoenholz. - The departure of Barret Zoph was contentious; Thinking Machines CEO Mira Murati stated she "parted ways" with him over "unethical conduct," while an internal OpenAI memo said the company did not share these concerns and had been in talks with him for weeks. - Mira Murati's startup, founded in February 2025, raised a staggering $2 billion seed round at a $12 billion valuation by July 2025, with backers including Andreessen Horowitz and Nvidia. - The name "Thinking Machines" is a nod to the original Thinking Machines Corporation, a pioneering supercomputer company from the 1980s and early 1990s that built some of the world's fastest computers before filing for bankruptcy in 1994. - This "boomerang" hiring is part of a larger trend dubbed the "AI lab revolving door," with top talent frequently moving between major players like OpenAI, Anthropic, Google, and well-funded startups, driven by compensation, mission alignment, and creative autonomy. - The intense competition for talent is localized in the Bay Area, which saw a 24% year-over-year increase in its AI-skilled tech workforce in 2025 and has the highest concentration of AI jobs in the U.S. - For top technical roles, compensation benchmarks are being set at record highs; reports indicate that salaries for some engineering positions at Thinking Machines Lab range from $450,000 to $500,000. - The movement of talent isn't just about money; it often reflects deep divisions within the AI community on the balance between the speed of commercial development and long-term AI safety research, a key factor for engineers choosing between labs.
