Alphabet Disbands AlphaFold Team to Scale Gemini AI

Google DeepMind has disbanded the dedicated team responsible for AlphaFold, the system that solved the 50 year protein folding problem. This structural shift reassigns researchers to the Gemini large language model and broader scientific initiatives. While the AlphaFold database of 200 million protein structures remains open, the specialized unit that built it no longer exists as a standalone entity.

This move marks a fundamental pivot in DeepMind's operational philosophy. For nearly a decade, the lab followed a strategy of pointing dedicated teams at specific grand challenges. AlphaFold was the pinnacle of this approach, providing the foundation for modern artificial intelligence in medicine and drug discovery. Now, the company is moving toward a platform centric model where Gemini serves as the core engine for various scientific domains.

DeepMind Research VP Pushmeet Kohli confirmed the strategy has evolved. Instead of isolated teams, the lab is now building Gemini powered systems designed to automate research and assist scientists across multiple fields. These include enzyme design, nuclear fusion, and genomics. Some staff have also transitioned to Isomorphic Labs, the Alphabet spinout focused on drug discovery.

The talent migration, however, has not been entirely internal. According to The Next Web, nearly a quarter of the original AlphaFold paper authors have left the company. Most notably, John Jumper, who shared the 2024 Nobel Prize, departed for Anthropic in June. He was followed by core researchers Jonas Adler and Alexander Pritzel.

This exodus is particularly timely given Anthropic's recent launch of Claude Science. This new workbench is specifically engineered for the biology and drug discovery work that AlphaFold pioneered. The loss of instrumental members has reportedly caused internal surprise at DeepMind, as a direct competitor now possesses the human capital to challenge Google's dominance in computational biology.

Alphabet is executing this reorganization while facing a volatile competitive landscape. Recent stock market news today shows the company's shares trading around $336.71, reflecting strong multiyear gains despite short term fluctuations. As reported by Yahoo Finance, the reallocation of talent suggests a tighter integration between Alphabet's core AI platforms and its long term scientific ambitions.

Beyond biology, the pressure to generalize AI is intensifying. While DeepMind pushes Gemini, other labs are shipping specialized models at a rapid pace. ZDNET notes that Microsoft's Cyber-1-Flash recently outperformed Gemini and OpenAI's GPT-5.6 on the CyberGym security benchmark. This suggests that while general models are the corporate goal, specialized performance remains the primary metric for practitioners.

The transition from AlphaFold to Gemini represents the industrialization of AI research. By treating AlphaFold as a solved problem, Alphabet is betting that a generalist model can synthesize knowledge across genomics and physics more efficiently than a dozen specialized teams. It is a high stakes gamble on the scalability of LLMs as scientific reasoning engines.

If Gemini fails to match the precision of the original AlphaFold team, Alphabet may find that it traded a Nobel winning specialty for a generalist tool that lacks the depth required for breakthrough science. The real test will be whether Gemini can move the needle on drug discovery as effectively as its predecessor.

Will the era of the dedicated AI research team be replaced entirely by foundational model tuning?

FAQ

Is the AlphaFold database still available? Yes, the database containing over 200 million protein structures remains open for researchers.

Where did the AlphaFold researchers go? Most were reassigned to Gemini or other DeepMind projects like nuclear fusion, though some joined Anthropic.

Why did Alphabet disband the team? The company is shifting from a one team per problem strategy to a broader approach centered on Gemini AI.

Who is John Jumper? A core AlphaFold researcher and 2024 Nobel Prize winner who recently moved to Anthropic.