Top U.S. professors are leaving universities in large numbers to join AI giants like OpenAI and Anthropic.

```

Top universities in the United States are facing a new wave of AI talent exodus.

As tech giants like OpenAI, Anthropic, and Meta continue to attract academic elites with generous resources, professors of computer science at several leading universities are accelerating their departures this year. Researchers remaining in academia warn that this trend not only threatens the development of open-source AI models, but also undermines the foundation of the United States in the AI field.

According to The Information, just since the beginning of this year, leading AI labs have recruited several renowned professors from top universities, and the pace of departures has noticeably sped up. Joey Gonzalez, a computer science professor at the University of California, Berkeley, noted that the number of faculty leaving has soared this year, and the resulting ripple effects are "far more severe than most people realize."

The most immediate concern is the outlook for open-source AI. Jennifer Chayes, Dean of Berkeley’s College of Computing, pointed out that top talent continuously flowing into commercial labs developing proprietary closed-source models will weaken the West’s competitiveness in open-source AI—a key front in the fight against Chinese models.

Wave of Departure: Several Professors Have Left This Year

According to The Information's summary, several university scholars have joined leading AI organizations so far this year. Stanford computer science professor Sanmi Koyejo joined Meta’s super-intelligence lab this month; Bo Li from the University of Illinois Urbana-Champaign also joined Meta in July; University of Virginia scholar Maissam Barkeshli joined Anthropic in July; University of Southern California professor Rahul Jain joined Google DeepMind in March on academic leave.

This wave of departures is not an isolated phenomenon. About fifteen years ago, with the advent of deep learning and autonomous driving software, AI professors began taking senior roles at companies like Google. Following the generative AI boom in recent years, this trend has accelerated significantly. The driving force remains unchanged: top-level AI research depends on massive, expensive computing resources, which most universities simply cannot provide.

Open Source Ecosystem Damaged: The Deeper Cost of Talent Loss

The impact of talent outflow on the open-source AI ecosystem is particularly concerning. Prior to his departure, Sanmi Koyejo supervised a Stanford doctoral student who developed the well-received Marin 32B open-source model. Whether Koyejo will continue teaching or take full leave after joining Meta remains unclear. Before joining DeepMind, Rahul Jain’s research focused on reinforcement learning for small-scale open-source AI models.

Even scholars not directly involved in open-source model development have invaluable value. Academic researchers in computer science have long published papers publicly, providing foundational support for other researchers to train and deploy AI models more efficiently and safely. Once these talents enter the closed commercial environment, the accumulation of public knowledge will slow down.

Virtuous Cycle Hard to Sustain

Joey Gonzalez described an ideal academic ecosystem loop: scholars enter industry to gain technical experience and unique insights, then return to the classroom to pass that experience on to students. However, he is pessimistic about whether this cycle can continue. He predicts that most colleagues who enter AI labs on academic leave will choose to stay in the industry rather than return to campus.

"If I get to use industry’s powerful research resources, it’s really hard to ask me to leave and go back to the university, doing research with graduate students who have extremely limited equipment," Gonzalez said.

This dilemma reflects the structural resource disadvantage universities face in the AI race. As the competition for computing power continues to escalate, the gap between academic institutions and commercial labs will only widen, and there is currently no clear answer on how to close it.

Risk Warning and DisclaimerThe market comes with risks, and investments need to be made with caution. This article does not constitute personal investment advice, nor does it consider the special investment objectives, financial circumstances, or needs of individual users. Users should consider whether any opinions, viewpoints, or conclusions in this article fit their specific circumstances. Investment decisions made accordingly are at one’s own risk. ```