Mountain View, California
Last year, for every three engineers who left Anthropic for Google DeepMind, thirty-seven went the other way. On June 19, 2026, the most accomplished of them all joined that trend.
John Jumper, who won the 2024 Nobel Prize in Chemistry for creating AlphaFold and was a vice president at Google DeepMind, announced he was leaving after nearly nine years to join Anthropic, the company behind Claude. His announcement came just one day after Noam Shazeer, co-lead of Google’s Gemini models and co-author of the influential “Attention Is All You Need” paper, said he was leaving for OpenAI. In the same week, Google lost the leaders behind its two most important AI breakthroughs.
This is more than a story about people changing jobs. It raises a bigger question: were the most important AI breakthroughs at Google the result of the company itself, or of the talented individuals who are now leaving?
What AlphaFold Actually Was — and Why It Matters Now
To see what Google DeepMind is losing, it helps to look at what John Jumper created. Protein folding, which means predicting a protein’s 3D shape from its amino acid sequence, was a major scientific challenge for 50 years. AlphaFold2 solved this problem. Since its launch, over two million scientists in 190 countries have used it to accelerate research on malaria vaccines, cancer treatments, and drug-resistant bacteria.
AlphaFold turned decades of expected progress in protein science into a database that anyone can access online. This is not an exaggeration; it is what the Nobel committee said when awarding the prize. John Jumper, born in 1985, became the youngest chemistry Nobel laureate in over 70 years.
Demis Hassabis, who shared the Nobel with Jumper, responded to the news in public and with grace. He said their work together “changed the world” and praised AlphaFold for showing what AI can do for science and medicine. However, Hassabis did not mention what DeepMind will do without the person most responsible for that success.
The AI Talent War Has a Clear Winner Right Now
The numbers in this AI talent war are clear. SignalFire’s 2025 State of Talent Report says engineers at DeepMind were almost 11 times more likely to leave for Anthropic than the other way around. This means hiring Jumper is not a surprise win, but part of a trend that has been growing for over a year.
What is causing this trend? Anthropic offers something that money alone cannot match: a clear scientific mission and a small, fast-moving team. Anthropic has kept 80% of its staff over two years, the best rate among top labs. Engineers at OpenAI are eight times more likely to leave for Anthropic than the other way around. These numbers show Anthropic is winning because of its purpose, not just pay.
For Google DeepMind, the bigger issue is not just one person leaving. When two top scientific leaders leave in the same week, each going to a main competitor, it sends a message about the company as a whole, not only about those individuals.
What Anthropic Gets — and What It Is Building Toward
Hiring Jumper fits with Anthropic’s growing focus on life sciences and computational biology. This is a planned move, not just a lucky hire. Anthropic had already created VirBench biology benchmarks, formed wet-lab partnerships with the Allen Institute and HHMI, and developed AI-for-science agent systems before bringing Jumper on board. He is joining a company that has already prepared the way, and now he is expected to help set the direction.
That goal has strong support from the top. Anthropic CEO Dario Amodei has written that AI-powered biology could achieve 50 to 100 years of scientific development in just one decade. By any standard, John Jumper is the person who has most clearly shown that this kind of rapid progress is possible.
The timing is significant. Anthropic is holding a science-focused event on June 30, and Jumper’s arrival puts him in a position to help shape what comes next. His exact role has not been announced. That lack of detail may be intentional. Anthropic is not simply placing a Nobel laureate into an existing role; it is letting him help design a new one.
The Nobel Prize AI Researcher Leaves Google for Anthropic — And the Coding Gap Widens
When a Nobel prize AI researcher leaves Google for Anthropic, it affects Google’s story in the business world. According to Bloomberg, DeepMind employees and leaders have said the company does not have a clear answer for enterprises searching for AI coding tools. In this area, Anthropic and OpenAI have made strong progress. Anthropic’s Claude Code has been a major driver of its recent revenue growth.
This is where the AI talent war turns into a business issue, not just a matter of reputation. Enterprise customers are not buying Nobel Prize awards; they are buying confidence that their chosen AI provider will still be a leader in two years. When top scientists keep leaving, that confidence goes with them.
Bloomberg also notes that Jumper was a key member of Google’s AI coding development team. This links his scientific background directly to the enterprise coding area, where Google DeepMind is already struggling to maintain its leadership. His deep knowledge of Google’s internal AI systems leaves with him.
The Question Google DeepMind Cannot Yet Answer.
This departure raises a big question: was AlphaFold the result of a repeatable process, or a unique achievement by a special team? The answer is important. If AlphaFold came from Google DeepMind’s systems and culture, the lab can create something similar again. But if it was mainly the work of John Jumper and his team, then Google is left with the name, not the know-how.
Anthropic is betting that it was the people, not just the process. In the AI talent race, the real prize is not market share but the few researchers with the right mix of scientific skill, engineering talent, and proven results to advance the discipline. John Jumper has already done this once.
What John Jumper does next matters not just to Anthropic’s investors or Google DeepMind’s leaders. It matters to every business leader who has based their AI plans on the idea that being big and established is a lasting advantage. In today’s AI talent war, that is no longer true. Scientists are making their choices clear, and the numbers are not close.
