Google is shifting its AI development toward centralization, with DeepMind increasingly becoming a subsidiary. Founder Demis Hassabis may soon leave the company, as operational leadership is moving to Koray Kavukcuoglu in the U.S. Hassabis was 'promoted' to chairman to ease his exit, after concerns about a potential stock drop following his role change. He plans to focus more on scientific work and his pharmaceutical startup, Isomorphic Labs. According to a former Google manager, the 'era of DeepMind as an independent actor is over,' and a current employee now refers to the lab as 'just another subdivision' of Google. Source: thedecoder
Google co-founder Sergey Brin is expected to take on a more influential role, especially as all Gemini-related AI development moves to the Bay Area. Brin had already returned to active development during Google's first 'red alert' over ChatGPT and was involved in the merger of Google Brain and DeepMind into 'Google DeepMind.' This merger also influenced the naming of the Gemini models, as the 'twin' teams from both organizations were meant to work together on a joint project. Tensions have existed for years, with reports indicating frustration from Hassabis and others over restructuring efforts. Source: thedecoder
Some experts view the restructuring as a strategic pivot toward Google's profitable cloud and infrastructure business, while others see it as a sign that Google is falling behind in the race for leading AI models. SemiAnalysis compares this to IBM's retreat from the PC business and Intel's shift away from bold investments, both of which were seen as mistakes. Meanwhile, Tim O'Reilly argues that Google is focusing on spreading AI technology rather than model leadership, using Thomas Edison and George Westinghouse as an example. Google's latest Flash models are optimized for efficiency while remaining useful. Source: thedecoder
