Jack Clark, co-founder of Anthropic, delivers a lecture at Oxford arguing that AI progress is accelerating beyond what most people acknowledge. Drawing on personal experience and observations from inside Anthropic, he describes how AI has transformed his daily life and work — from personal advice to research acceleration. He outlines how Anthropic itself is changing: most code is now written by Claude, employees are shifting to verification and management roles, and a single researcher can direct 9 AI agents. Clark makes bold predictions including autonomous companies, AI-Nobel-Prize discoveries by 2027, and recursive self-improvement by 2028. He argues society faces a choice between actively shaping this future or being caught in reactive denial. The issue also includes a fictional short story imagining a family whose terminally ill daughter is cryogenically preserved and later revived after an AI-driven intelligence explosion enables medical breakthroughs.
Sort: