πŸ“° Key Summary

Barret Zoph has switched jobs again, becoming the latest example of high-level personnel churn in the AI industry. He spent two years at OpenAI before leaving in October 2024 to co-found startup Thinking Machines alongside fellow OpenAI departee Mira Murati. This January, Zoph and co-founder Luke Metz abruptly and dramatically left Thinking Machines to return to OpenAI, and it later came out that he’d actually been fired. Back at OpenAI, he was put in charge of the enterprise sales division β€” but that gig didn’t last long either, ending just five months later when he left again this June. Now his next stop is confirmed: VP of Research at Google, a company he’d worked at before. A Google spokesperson told The Wall Street Journal they’re looking forward to Zoph bringing his expertise in reinforcement learning (RL) and post-training back to help the Gemini team. The article also notes that turnover among AI industry leadership is generally high β€” especially at OpenAI. Even as the company prepares for an IPO and remains one of the most influential in tech, it’s lost a string of key people over the past eight months, including its COO and, more recently, a senior data center executive, leaving outsiders puzzled about what’s going on internally.

πŸ’¬ JudyAI Lab Take

Google’s research division has a new VP, and the pick is none other than Barret Zoph, who’s bounced between OpenAI and Thinking Machines over the past year β€” another sign of just how fluid top-tier AI talent movement has become.

Look at Zoph’s path: co-founding a startup, getting fired, returning to OpenAI to run sales, leaving again five months later, and finally landing at Google to lead research β€” all within a little over a year. That pace reflects something real happening in the AI talent market right now: senior researchers are switching roles far faster than they used to, as companies compete for talent while those same people are still figuring out where they fit. Worth noting for AI builders: Google specifically wants Zoph’s reinforcement learning and post-training chops to strengthen Gemini, which tells you those two capabilities remain a core battleground across the industry. And OpenAI losing its COO plus a senior data center exec within eight months shows that even the category leader isn’t immune to questions about internal stability.

If you’re tracking AI industry moves, it’s worth keeping an eye on research-lead changes at these companies β€” they often signal a shift in technical direction well before any product launch does.

πŸ“… Source Info

πŸ”— Further Reading