Ep 88: Unpacking DeepMind's Quest for SuperIntelligence with Demis Hassabis' Biographer

The AI race is an inevitable consequence of powerful technology, driving multiple labs to compete for supremacy despite early hopes for a unified, safe development path. Sebastian Mallaby, author of *The Infinity Machine*, details how DeepMind co-founder Demis Hassabis initially envisioned a singular, controlled approach to AGI, only to face a crowded field that necessitated a shift toward viewing AI safety as a collective action problem requiring government intervention. While Google’s DeepMind has consistently pushed scientific boundaries, including the Nobel-winning AlphaFold, it has struggled to match the product-focused agility of startups like OpenAI and Anthropic. The narrative surrounding these leaders is defined by intense personal histories, such as the failed attempts to spin DeepMind out of Google and the shifting alliances between figures like Hassabis, Sam Altman, and Elon Musk, illustrating the complex interplay between scientific ambition and commercial reality.