AI talent war: Nobel winner John Jumper leaves Google for Anthropic, Noam Shazeer returns to OpenAI
The AI talent war just got louder. AI means “artificial intelligence” — computer systems that can learn and answer like a smart person. Two of the world’s best AI scientists are now changing sides. John Jumper, who won a Nobel Prize, is leaving Google DeepMind to join a company called Anthropic. And Noam Shazeer, who used to help lead Google’s Gemini AI, is going back to OpenAI. Both moves happened within days of each other in June 2026. They show us how hard the big AI companies are fighting to win.
An “AI lab” is a company that builds powerful AI systems. The biggest ones are Google DeepMind, OpenAI (the maker of ChatGPT), and Anthropic (the maker of Claude). All three want the smartest people. So they keep taking each other’s stars. When one company hires a star away from another, people call it “poaching.”
Who is John Jumper?
John Jumper is one of the most respected AI scientists alive. He led the team that built a tool called AlphaFold at Google DeepMind. AlphaFold is an AI tool that guesses the shape of proteins. Proteins are tiny building blocks inside our bodies. Knowing their shape helps scientists understand diseases. It also helps them make new medicines faster.
Before AlphaFold, finding one protein’s shape could take years in a lab. AlphaFold has now found the shapes of more than 200 million proteins. That saved years of work in medicine and biology.
For this work, Jumper won the 2024 Nobel Prize in Chemistry. A Nobel Prize is the world’s most famous science award. Winning one puts you in a very small group of people. He shared the prize with Demis Hassabis, the CEO (the top boss) of Google DeepMind. Jumper worked at Google DeepMind for almost nine years. Now he is moving to Anthropic, the company that makes the Claude AI assistant. The news came out on June 19, 2026.
Hassabis was kind about the exit. He said AlphaFold had “changed the world.” He also praised their “extraordinary partnership.”
Who is Noam Shazeer?
Noam Shazeer is a legend among AI experts. In 2017 he helped write a research paper called “Attention Is All You Need.” A research paper is a written report that shares a new idea. Many people call it the most important AI paper ever. It introduced the “transformer.” The transformer is the basic design behind almost every modern AI chatbot, including ChatGPT and Gemini. So in a real way, Shazeer helped invent the technology that this whole AI boom runs on.
His career has gone in a loop. He joined Google in 2000. He even made its spell checker better. He left in 2021 to start his own chatbot company, Character.AI. In 2024, Google brought him back through a $2.7 billion deal. That deal brought Shazeer, his co-founder Daniel De Freitas, and part of their team back to Google. He then became a Vice President. He co-led the Gemini AI models with Jeff Dean and Oriol Vinyals.
Now, in June 2026, he is leaving Google again. This time he is joining OpenAI. He said it was a hard choice. “It was a difficult decision to move on,” he said. There is a twist. Shazeer first came back to Google to help fix its AI models, which had fallen behind OpenAI and Anthropic. Now he is joining the very rival he was hired to beat.
Key facts at a glance
| Detail | John Jumper | Noam Shazeer |
|---|---|---|
| Leaving | Google DeepMind | |
| Joining | Anthropic | OpenAI |
| Known for | AlphaFold protein AI | “Attention Is All You Need” paper |
| Big credential | 2024 Nobel Prize in Chemistry | Co-creator of the transformer |
| Time at Google | Nearly 9 years | Returned via $2.7 billion deal in 2024 |
| News date | June 19, 2026 | June 2026 (days earlier) |
Why do top AI researchers matter so much?
AI is not built by big teams alone. A small number of brilliant minds often set the path. One great researcher can find the idea that makes a model much smarter. That is why labs treat these people like star athletes.
These people are rare. There may be only a few hundred who can lead work at the very top. When one of them moves, they take their knowledge and methods with them. They often take a small team too. So losing one star hurts twice. The rival gets stronger, and you get weaker at the same time.
- They choose what to build next and how to build it.
- They pull in other top people who want to work with them.
- They carry hard-won lessons that are not written in any book.
- They can speed up a lab by months, which is huge in this race.
What these moves tell us about the AI race
These two exits are not random. They show a clear pattern. Google DeepMind is losing big names. Besides Jumper and Shazeer, David Silver also left recently. Silver was a key researcher behind AlphaGo and AlphaZero. Those are AI systems that beat the world’s best human players at the board game Go. He left to start his own company.
So the talent is moving in one direction. It is going away from Google and toward Anthropic and OpenAI. Some insiders even worry that Google’s next model, said to be called Gemini 3.5 Pro, may struggle against rivals. When your best people leave for the competition, it raises questions about morale and direction.
This also shows how fierce the fight has become. OpenAI is reportedly getting ready for an IPO. An IPO is the first time a company sells its shares to the public, so anyone can buy a small piece of it. Anthropic is growing fast too. Both have a lot of money, and they are spending it to win the best brains. The fact that a Nobel winner would switch labs tells you how high the stakes are.
Why it matters (especially for India and founders)
For Indian founders and students, there are real lessons here. A founder is a person who starts a company. The first lesson is simple: talent is the real asset. The most valuable thing a tech company owns is not its office or even its code. It is the people. If a Nobel winner can move and shake a whole company, think about what your best engineer means to your startup.
The second lesson is that keeping people matters as much as hiring them. Google paid $2.7 billion to bring Shazeer back. Then it lost him anyway. Money alone does not keep great people. They stay for a mission, for freedom, and for the chance to make an impact. Founders should think hard about company culture, not just pay.
The third lesson is good news for AI builders everywhere, including in India. As the labs fight, AI tools get better and cheaper fast. Indian startups can build on top of these models without owning a giant lab of their own. The competition you are reading about is, in the end, helping you.
Frequently asked questions
Who is John Jumper and why is his move big news?
John Jumper led AlphaFold at Google DeepMind. He won the 2024 Nobel Prize in Chemistry. His move to Anthropic is big because Nobel-level scientists almost never switch labs. It shows how hard rivals are fighting for top talent.
Why is Noam Shazeer leaving Google for OpenAI?
Shazeer called it “a difficult decision to move on.” He co-led Google’s Gemini AI after coming back in a $2.7 billion deal in 2024. Now he is joining OpenAI, a direct rival, in June 2026.
What do these moves mean for the AI race?
They suggest Google DeepMind is losing key people to Anthropic and OpenAI. With other exits like David Silver too, it points to tough competition. It may also mean pressure on Google’s next models.
The takeaway
The AI talent war is now a big story of 2026. In one week, a Nobel winner and a co-inventor of the transformer both switched labs. The message is clear. The race to build the smartest AI is really a race to win the smartest people. So watch where the top researchers go. That often tells you which lab will lead next.
Sources: The Decoder on John Jumper joining Anthropic and The Decoder on Noam Shazeer joining OpenAI.
