Bob McGrew on No Priors: Superintelligence Needs Physical AI and Why Hard Takeoff Is Wrong
saranormous · x · 2026-11-20
No Priors interviews Arda cofounder Bob McGrew, former Chief Research Officer at OpenAI and early Palantir exec. Key threads:
- Why superintelligence requires physical AI, and why the hard-takeoff argument is wrong
- Labs vs startups in robotics and biology; open-world vs closed-world robotics; humanoids vs robotic work cells
- Who needs flexible automation: defense, consumer electronics, data centers; building with Anduril cut delivery from days to 90 minutes
- Palantir's problem-first approach vs OpenAI's technology-first; how Forward Deployed Engineering really works
- Why general robotics models suddenly work; on-device models, speed, and safety
- Self-replicating factories, Von Neumann probes, and his answer to the Fermi paradox
Also covers dropping out of Stanford to join OpenAI as a nonprofit, and why America designs products but doesn't build them.
Related event: Ex-OpenAI Research Chief Bob McGrew: Superintelligence Needs Physical AI(3 posts)→
More from AGI Musings
- The 'Trojan Horse externality': AI slop looks exactly like real production — round · 2026-11-20
- ludwigABAP digs up 20 old 'collapse' essays and cringes at his own AGI doom predictions — ludwigABAP · 2026-11-20
- 'A year can change an industry': Anthropic IPO and Jensen Huang's exit one year after Opus 4.5 — ColleenMBrady · 2026-11-20
- Ex-Anthropic/OpenAI researcher bets 80% of cognitive labor automated by end of 2028 — rohanpaul_ai · 2026-11-20
- Herbie Bradley: preventing model weight theft via SL5 security imposes huge costs — herbiebradley · 2026-11-20
- Kevin Xu shares what zero-sum AI looks like, Wall Street edition — kevinsxu · 2026-11-20