In less than a week, Google lost two of its most important AI researchers. To two of its direct competitors.
On June 22, Noam Shazeer announced on X that he was moving to OpenAI. A few days earlier, John Jumper had announced his own move to Anthropic.
Alphabet's stock plunged 5-6%, with an estimated market cap loss of around $250 billion.
Who the Two Researchers Are
Noam Shazeer isn't just a heavyweight name on Google's roster — he's one of the co-authors of the "Attention Is All You Need" paper (2017), the work that introduced the Transformer architecture. That research is the technological foundation on which almost all of today's large language models are built: GPT, Claude, Gemini, LLaMA, and others.
Google had brought him back to the company in 2024 through the $2.7 billion acquisition of Character.AI, installing him as co-lead of the Gemini project. Less than two years later, he's leaving for OpenAI.
John Jumper led the AlphaFold project at Google DeepMind, the AI system that solved one of the hardest problems in computational biology: predicting the three-dimensional structure of proteins. For this work, he shared the 2024 Nobel Prize in Chemistry. He's now taking his expertise to Anthropic.
The Signal Behind the Departures
It's not just about résumés. It's what the departures indicate about the internal state of Google AI.
Bloomberg reported that DeepMind researchers had internally raised concerns about a specific issue: the lack of a clear product for companies building AI coding tools — an area where Anthropic (with Claude Code) and OpenAI (with Copilot) have gained significant ground.
Google has a huge advantage in distribution — every time you open Chrome or search on Google, you can tap into Gemini. But in the professional developer tools segment, that competitive edge has shrunk. And that's exactly where the two departing researchers are headed.
The Bigger Picture
Google is under pressure from two opposite directions:
From above: AI Overviews are changing the search experience and putting pressure on the advertising model — people who get their answer from AI click on fewer ads.
From below: competitors are attracting top-tier talent, including researchers Google itself trained or brought back at great expense.
The market responded by punishing the stock. The question analysts are asking is whether Shazeer's and Jumper's departures are an isolated signal or part of a deeper trend in Google AI's competitive positioning.
For those following the AI sector, this week carried the weight of a turning point. Not because of the products still to come — those will show up later — but because of the narrative taking shape: whoever manages to attract (and retain) the best researchers has a structural advantage that isn't measured only in model benchmarks.