Demis Hassabis mentions that Google DeepMind needed to revisit its startup origins following the merger with Brain.

      In summary: Demis Hassabis, during an early April 2026 appearance on the 20VC podcast with Harry Stebbings, elaborated on how Google DeepMind has sped up its progress over the last two to three years by integrating the computing power of Google Brain with DeepMind’s research ethos and reverting to a “startup or entrepreneurial” approach. He also mentioned that he manages Isomorphic Labs, the group’s AI-focused pharmaceutical spinoff, during what he termed a “second workday” starting around 10pm, in anticipation of upcoming human trials in oncology later this year.

      Gathering the elements

      The official merger of DeepMind and Google Brain was finalized in 2023. Hassabis indicated that the time since then has been marked by intentional acceleration: aligning talent “from across the company, sort of pushing in one direction,” gaining substantial access to computing infrastructure that DeepMind had previously lacked, and fostering what he called a “relentless sort of focus and pace.” He explained that the transformation demanded both cultural and structural shifts, necessitating the organization to “return to our startup or entrepreneurial roots, be scrappy, be quicker, and deliver things rapidly.” He described the current competitive landscape as “ferocious.” Long-time employees with 20 to 30 years in the industry have told him that it is “the most intense environment they’ve ever witnessed, perhaps in the technology sector.”

      Hassabis noted that he communicates with Sundar Pichai, the CEO of Alphabet, “every day,” underscoring the degree to which Google DeepMind plays a central role in the operational strategy of Alphabet’s products and research. This close relationship is backed by a substantial financial commitment. Google’s expansion of computational resources, partly developed through partnerships for custom chips with firms like Broadcom, is critical to that positioning: Alphabet invested $91.4 billion in capital expenditure in 2025 and has forecasted between $175 billion and $185 billion for 2026, nearly doubling the previous year’s investments, with supply constraints cited as the primary limiting factor rather than capital access.

      The 90% assertion

      One of Hassabis’s more assertive remarks on the podcast addressed DeepMind’s role in the evolution of AI. He stated that around 90% of the innovations forming the foundation of the modern AI sector originated from either Google Brain, Google Research, or DeepMind. This assertion aligns broadly with documented academic accomplishments, including the transformer architecture introduced by Google Brain in 2017, initial research on reinforcement learning from human feedback, and deep reinforcement learning methods created at DeepMind. The 2024 Nobel Prize in Chemistry awarded to Hassabis and John Jumper, in collaboration with David Baker, for the AlphaFold protein-folding solution stands as the most officially recognized of these successes. Whether the 90% claim accurately represents a proportion is subject to interpretation, and the industry has diversified significantly since those initial studies. The framing serves as both a positioning statement and a historical assertion.

      The operational impact of this legacy is demonstrated by a significantly accelerated product release schedule. Google’s open-weight model program, most recently including Gemma 4, now releases models developed from the same research and training frameworks as Gemini 3, bridging a gap that once existed between cutting-edge research and open-source contributions. By the end of the fourth quarter of 2025, Gemini had acquired about 750 million monthly active users, and reports indicate that the launch of Gemini 3 in November of that year prompted an urgent response within OpenAI.

      The second workday

      In addition to leading Google DeepMind, Hassabis oversees Isomorphic Labs, an AI-driven pharmaceutical spinoff that DeepMind created in 2021. He explained his work routine during the 20VC discussion: a first workday at DeepMind followed by a “second workday” starting around 10pm focused on Isomorphic’s drug discovery efforts. This dual role reflects his belief that leveraging AI for drug discovery is both his most significant long-term objective and a task necessitating ongoing personal engagement rather than delegation.

      Isomorphic secured $600 million in funding in April 2025 and has entered partnership deals with Eli Lilly and Novartis, with combined milestone payments potentially reaching $3 billion. In February 2026, the company introduced IsoDDE, a drug design tool that Isomorphic claims enhances AlphaFold 3’s accuracy in generating drug candidates by two-fold. Human clinical trials in oncology are anticipated later in 2026. The competitive landscape in AI-driven drug discovery is intensifying: Anthropic’s acquisition of Coefficient Bio for approximately $400 million in April 2026—a stealth startup founded by former Genentech computational biology researchers—indicates that general-purpose AI companies are increasingly viewing pharmaceutical discovery as a significant product area rather than just an illustration of model performance.

      The competitive context

      The discussion on the 20VC podcast, similar to Sebastian Mallaby’s biography of Hassabis, "The