Google DeepMind shifting prioritiesTop Google DeepMind executive says ‘strategy has evolved’Google has seen some high-profile exits to competitors in recent monthsGoogle DeepMind has dismantled the research team behind AlphaFold, its Nobel Prize-winning AI system that has been designed for predicting protein structures.
The restructuring marks a major strategic shift for the laboratory as it realigns its focus toward large language models (LLMs) and general scientific automation.According to an analysis of recent job moves and company insiders undertaken by The Financial Times, the majority of the original AlphaFold paper authors have been reassigned internally or departed the company.
The accomplishment earned DeepMind CEO Demis Hassabis and vice president/engineering fellow John Jumper the Nobel Prize in Chemistry in 2024.
However, key leaders behind the project have recently departed:Jumper, co-recipient of the 2024 Nobel Prize, recently left DeepMind to join competitor Anthropic.
Core AlphaFold researchers Jonas Adler and Alexander Pritzel have also exited and joined Anthropic.
Google DeepMind shifting priorities
Top Google DeepMind executive says ‘strategy has evolved’
Google has seen some high-profile exits to competitors in recent months
Google DeepMind has dismantled the research team behind AlphaFold, its Nobel Prize-winning AI system that has been designed for predicting protein structures. The restructuring marks a major strategic shift for the laboratory as it realigns its focus toward large language models (LLMs) and general scientific automation.According to an analysis of recent job moves and company insiders undertaken by The Financial Times, the majority of the original AlphaFold paper authors have been reassigned internally or departed the company. Almost a quarter of the full-time DeepMind authors who built the original AlphaFold papers have left Google altogether, the report said.Google DeepMind confirmed to the publication that remaining researchers have transitioned to other internal initiatives, including projects built around Google's flagship Gemini model, nuclear fusion, genomics and enzyme design while others moved to Alphabet's commercial drug-discovery spin-off, Isomorphic Labs.The dissolution of the AlphaFold unit points to how the rise of LLMs has reshaped priorities across top AI laboratories. Pushmeet Kohli, vice president of research at Google DeepMind and head of its AI for Science team, explained that the lab is moving away from assembling isolated teams around single, long-term scientific targets.“Our strategy over the last nine years has been to focus on grand challenges... a concrete goal every project is focused on.The strategy has evolved,” Kohli said.Instead of focusing on standalone problems like protein folding, DeepMind is reportedly prioritising Gemini-powered systems capable of assisting scientists across various fields and eventually automating parts of the research process, while competing directly with rivals like OpenAI and Anthropic.Launched in 2018, AlphaFold revolutionised biological research by solving a 50-year-old challenge in structural biology. The accomplishment earned DeepMind CEO Demis Hassabis and vice president/engineering fellow John Jumper the Nobel Prize in Chemistry in 2024. However, key leaders behind the project have recently departed:Jumper, co-recipient of the 2024 Nobel Prize, recently left DeepMind to join competitor Anthropic. Core AlphaFold researchers Jonas Adler and Alexander Pritzel have also exited and joined Anthropic.
