A looming intelligence explosion could make ‘everyone’s job obsolete’ — and it may arrive sooner than you think. (34→3 minutes) Krystal Ball and Saagar Enjeti sit down with Daniel Kokotello (ex-OpenAI, AI 2027) to map a near-term future where machines train machines, accelerating recursive self-improvement and risking rapid automation of white-collar and physical work. Learn why major labs are aiming for superintelligence, how current misalignment (sycophancy, reward hacking, Grok/Bing incidents) foreshadows bigger failures, and why concentration of power, grid strain, and economic upheaval follow. What you’ll take away: clear timelines (2028–2030 risk window), the twin dangers of loss of control and concentrated power, and three practical steps—wake up, demand transparency, and push coordinated regulation and protections. Listen now to get the key ideas in minutes.