John Jumper, the Nobel Prize-winning co-creator of AlphaFold, is leaving Google DeepMind after nearly nine years to join Anthropic. His departure, announced on social media, removes one of the most influential scientists in biology from the tech giant and hands a major recruitment victory to the AI startup, intensifying an already fierce competition for elite research talent.
Jumper shared the 2024 Nobel Prize in Chemistry with DeepMind chief executive Demis Hassabis for AlphaFold, an AI system that predicts the three-dimensional structures of proteins with remarkable accuracy. The technology has transformed biological research by mapping more than 200 million protein structures, work that previously took years or decades in a lab. He described Google DeepMind as a "special place" and said he would take a short break before starting at Anthropic. His new role has not been disclosed.
AlphaFold's impact on scientific discovery
Unlike general-purpose AI assistants, AlphaFold was built to solve a single, hard problem: predicting how proteins fold into complex shapes. A protein's shape determines its function, so accurate predictions open direct paths to understanding diseases and designing new drugs. The AI for Science & Research community has adopted the system at scale. The freely available AlphaFold Protein Structure Database now holds predictions for hundreds of millions of proteins, making it one of the largest scientific resources ever built with artificial intelligence.
According to the journal Nature, AlphaFold has become one of the most influential AI systems developed for science, with research groups worldwide using it to accelerate work in medicine, agriculture and biotechnology.
A talent war reshaping the industry
Jumper's move follows the recent exit of another senior Google AI researcher, Noam Shazeer, who left to join OpenAI. The back-to-back departures underline how a small group of scientists now holds outsized sway over the direction of artificial intelligence. Companies are pouring billions into computing infrastructure, but experienced frontier researchers are an even scarcer resource.
"There is so much demand for limited AI research talent," D.A. Davidson analyst Gil Luria told Reuters. He added that startups can often offer researchers less bureaucracy and a stronger focus on developing advanced AI systems.
What Jumper brings to Anthropic
Anthropic, founded in 2021 by former OpenAI researchers, has grown quickly through its Claude family of models and its public emphasis on AI safety. The company has secured billions in investment from Amazon and Google. Jumper's arrival signals that Anthropic plans to deepen its scientific research capabilities, not just compete in consumer AI. The company is expected to host a science-focused event later this month.
For Google DeepMind, the loss of another internationally recognised scientist comes as competition accelerates to build increasingly capable systems. For Anthropic, it represents one of the industry's biggest recruiting wins-and a sign that the next phase of the AI race may be decided as much by talent as by technology.
Why this matters for researchers in science
The movement of scientists like Jumper from large tech firms to specialised AI labs signals where the next wave of research tools will likely be built. When Nobel-level talent chooses a smaller, safety-focused company, it suggests that breakthroughs in drug discovery and protein modeling may increasingly come from outside the traditional tech giants. For researchers who depend on these tools, the shift could mean faster, more open access to next-generation models-or a more fragmented landscape of competing platforms. Tracking where the key people go is now a practical way to anticipate where the most useful scientific instruments will emerge.
Your membership also unlocks:
