{"text":[[{"start":8.78,"text":"Google DeepMind has dismantled the team behind AlphaFold, its Nobel Prize-winning AI system for predicting protein structures, as the company overhauls the research strategy that made it one of the world’s leading AI laboratories."}],[{"start":22.28,"text":"The majority of the original authors of the AlphaFold papers have been reassigned over the past year, according to an FT analysis of recent job moves and people familiar with the matter."}],[{"start":33.02,"text":"DeepMind confirmed the staffers have moved internally to projects built around Google’s Gemini large language model, as well as areas including enzyme design, nuclear fusion and genomics."}],[{"start":44.6,"text":"Others have relocated to Alphabet’s drug-discovery company Isomorphic Labs. Nearly a quarter of the full-time Google DeepMind authors of the original AlphaFold papers have left the company altogether."}],[{"start":55.92,"text":"The reshuffle reflects a significant change in DeepMind’s scientific strategy since the rise of large language models."}],[{"start":61.76,"text":"“Our strategy over the last nine years has been to focus on grand challenges . . . a concrete goal every project is focused on,” said Pushmeet Kohli, vice-president of research at Google DeepMind, where he founded and leads the AI for Science team. “The strategy has evolved.”"}],[{"start":78.56,"text":"He said rather than simply organising researchers around single scientific problems such as protein folding, the AI lab is now also focused on building Gemini-powered systems that can assist scientists — and eventually automate parts of the scientific process — while competing with OpenAI and Anthropic to develop frontier AI agents."}],[{"start":97.6,"text":"AlphaFold, which DeepMind began developing in 2018, transformed biological research by using AI to predict the three-dimensional structure of proteins, a breakthrough that earned John Jumper, vice-president, engineering fellow, and DeepMind chief executive Demis Hassabis, a Nobel Prize in Chemistry in 2024."}],[{"start":116.76,"text":"The project became the defining example of DeepMind’s strategy of pursuing long-term scientific breakthroughs through dedicated research teams."}],[{"start":125.12,"text":"Earlier this year, Jumper and fellow AlphaFold researcher Jonas Adler moved to Google’s internal “Code Strike” team, a group assembled to improve the company’s AI coding capabilities as it seeks to catch Anthropic and OpenAI."}],[{"start":137.76,"text":"Last month, Jumper announced he was leaving for Anthropic, where he will be joined by Adler and AlphaFold colleague Alexander Pritzel."}],[{"start":145.28,"text":"A DeepMind employee, who asked not to be named, said all three had been “instrumental, important, core members” of the company and that their departures had sparked surprise internally."}],[{"start":156.32,"text":"DeepMind said the movement of researchers reflected its evolving scientific priorities."}],[{"start":161.8,"text":"“We are incredibly proud of our scientific heritage and the global impact of AlphaFold, which inspired the launch of Isomorphic Labs with the mission to reimagine drug discovery,” the company said."}],[{"start":173.2,"text":"Isomorphic Labs was spun out of DeepMind in 2021 to commercialise AlphaFold’s scientific breakthroughs. It has since struck partnerships with pharmaceutical companies including Novartis and Eli Lilly as it seeks to use AI to speed drug discovery."}],[{"start":188.08,"text":"The reorganisation also reflects a broader shift across the AI industry, as frontier labs increasingly view large language models not simply as chatbots but as tools capable of accelerating scientific research. Anthropic this year launched Claude Science, a version of its AI assistant designed to support research in biology and drug discovery."}],[{"start":208.56,"text":"Competition for the small pool of researchers capable of building frontier AI systems has intensified alongside that strategic shift, with companies aggressively recruiting top scientists and engineers."}],[{"start":219.8,"text":"“[AI researchers] want to work for a lab that is at the frontier and [not] have to play a catch-up game,” Mark Chen, OpenAI’s head of research, told the FT earlier this year."}],[{"start":229.56,"text":"“It’s much more exciting for a researcher to just be at the frontier, trying to push the next paradigm, rather than all this pressure to replicate some model that is close to the frontier,” he added."}],[{"start":244.36,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1785308730_7033.mp3"}