Google DeepMind reassigns AlphaFold team members to Gemini projects, some leave company

AlphaFold Team Reassigned

Google DeepMind has disbanded the team responsible for AlphaFold, an AI program that accurately predicts protein structures. Most key members and original authors of AlphaFold papers have been reassigned internally, primarily to projects focused on Gemini, Google's AI model. A few team members have also left the company entirely.

AlphaFold's Impact and Achievements

AlphaFold, developed by DeepMind starting in 2018, was recognized in 2020 for solving the 50-year-old "protein folding problem." This AI can predict three-dimensional protein structures from amino acid sequences in minutes, a process that previously took years. Its applications include accelerating drug discovery, vaccine development, and understanding neurodegenerative diseases.

In 2021, DeepMind published AlphaFold's methodology and the structure predictions for the entire human proteome, making over 200 million protein structure predictions freely available to researchers through the AlphaFold Protein Structure Database. In 2024, DeepMind CEO Demis Hassabis and John Jumper, a key scientist on the project, received the Nobel Prize in Chemistry for their work on AlphaFold.

Personnel Changes and Strategic Shift

John Jumper, a VP and engineering fellow who was instrumental in AlphaFold, announced his departure in June to join Anthropic, with some colleagues also moving to the same company. DeepMind confirmed that other staff members were moved to Gemini-focused projects, and some were reassigned to Isomorphic Labs, Alphabet's drug-discovery company and a DeepMind spin-off.

This restructuring reflects Google's broader strategy to allocate more resources towards the development of Gemini, indicating a prioritization of this AI initiative within the company.