Yudkowsky's 2008 takeoff scenarios resurface as AI-advances-AI takes hold
jd_pressman · x · 2026-10-16
Rob Bensinger notes Yudkowsky explicitly ruled in 2008 that we might build AGI without deep new insights, and we're moving from scenario 2 (programmer-driven) to scenario 1 (AI advances AI); 2013-Eliezer's median takeoff was 2035. jdpressman argues reality is a mix of both scenarios, and scenario 1's ending isn't ruled out.
More from AGI Musings
- Ben Affleck calls AI 'just matrix multiplication'; Grady Booch echoes the skepticism — Grady_Booch · 2026-10-16
- Everyone has a precise AGI definition—until a model gets close to meeting it — VraserX · 2026-10-16
- Viral take: If AI gives us time back, its greatest gift is revealing what we mistook for the cost of living — r0ck3t23 · 2026-10-16
- "AI automation = slop, AI augmentation = super intelligence" — the line CEOs should see — jacob_posel · 2026-10-16
- Reddit take: Building a successor AI species would itself be humanity's purpose fulfilled — Flashy-Disk311 · 2026-10-16
- When Simple Rules Beat AI: A Frozen Ride-Hailing State Machine at Dulles — shashib · 2026-10-16