Google Reshuffles Nobel-Winning DeepMind AI Team Amid Shift in Research Strategy

A significant change is underway at Google’s DeepMind, a leading artificial intelligence laboratory. According to the Financial Times, most of the original authors behind the AlphaFold papers have been reassigned or left the company over the past year. This move marks a major shift in DeepMind’s scientific strategy following the advent of large language models.

The reorganization is part of Google’s broader effort to adapt its research approach and focus on developing systems that can help scientists, potentially automating parts of the process. The company confirmed that employees have moved to projects related to Gemini, its large language model, as well as areas like enzyme design, nuclear fusion, and genomics.

Almost a quarter of the full-time Google DeepMind authors who worked on the original AlphaFold papers have left the company altogether. This significant turnover is attributed to the changing landscape in AI research, with companies competing to build frontier AI agents. The shift in strategy has also led some researchers to join rival firms like OpenAI and Anthropic.

Pushmeet Kohli, vice-president of research at Google DeepMind, acknowledged that the company’s approach has evolved over time. ‘Our strategy over the last nine years has been to focus on grand challenges,’ he said. However, with the advent of large language models, DeepMind is now concentrating on developing Gemini-powered systems.

The news comes weeks after John Jumper, one of the Nobel Prize-winning scientists behind AlphaFold, announced his departure from Google to join Anthropic. This move highlights the intense competition in AI research and development.

In related news, recent studies have shown that conversational AI is being increasingly used by workers in various roles, including auto technicians and industrial mechanics. The company’s research found that workplace AI now touches 68% of jobs, representing 90% of employment in the U.S.