Google Merges DeepMind, Brain Research Units as AI Race Intensifies

Alphabet Inc.’s Google has consolidated its artificial intelligence research groups into one unit, the company’s latest move to keep from falling behind in the AI race.

The change folds the Brain team from Google Research and Alphabet’s DeepMind into one team, Chief Executive Officer Sundar Pichai said in a blog post on Thursday. “Combining all this talent into one focused team, backed by the computational resources of Google, will significantly accelerate our progress in AI,” Pichai said. Demis Hassabis will lead the group as CEO of DeepMind.

Jeff Dean, the executive who had been leading Google’s artificial intelligence and research efforts, will move out of management as part of the reorganization, according to two people familiar with the matter, who asked not to be identified because they weren’t authorized to discuss internal issues. In his new role as Google’s chief scientist, Dean will be working with both Google Research and DeepMind to develop new, more capable AI systems, and won't be overseeing large teams, the people said. He will report directly to Pichai.

Alphabet’s DeepMind, based in London, has long been known as the Google parent company unit that regularly introduced artificial intelligence breakthroughs, including its work on AlphaFold, technology that can predict the shape of proteins, as well as AlphaGo, software that taught itself to play the strategy game Go better than any human on earth. Internally, the unit has typically been seen as a group that works on artificial intelligence concepts that may not have direct applications in Google products.

Google Research, meanwhile, was responsible for “transformer” technology, key building blocks for large language models. That technology powers the current crop of chatbots, including Google’s Bard and OpenAI Inc.’s ChatGPT.