Demis Hassabis Wants to Build AGI. Should We Trust Him?

This episode of The Next Big Idea explores the life and motivations of Demis Hassabis, the visionary behind DeepMind and a central figure in the race to build artificial general intelligence (AGI). Journalist Sebastian Malaby, author of The Infinity Machine, shares insights from over 30 hours of interviews with Hassabis and more than 100 people in his orbit. The conversation delves into Hassabis’ improbable journey from a London chess prodigy to a Nobel Prize-winning neuroscientist and AI pioneer, highlighting his unique blend of compassion from his Christian mother and fierce competitiveness from his father. Malaby frames Hassabis as a man driven by a spiritual quest to understand the universe through science, yet one who grapples with the paradox of being both great and good. The episode examines his pivotal role in uniting deep learning and reinforcement learning, his breakthrough with AlphaFold, and his surprising lag in developing large language models—despite his early conviction that AGI requires embodied intelligence. As OpenAI surged ahead with ChatGPT, Hassabis was forced to confront the tension between his safety-first principles and the pressure to compete. The episode concludes with a meditation on whether one can be both a transformative genius and a moral actor in the age of superintelligence, ultimately suggesting that Hassabis’ pragmatism, humility, and commitment to open science make him a rare beacon of hope in the AI race.