PauseAI vs Andy Masley: does superintelligent AI risk require outlandish beliefs?
birchlse · x · 2026-11-12
PauseAI UK quote-rebutted Andy Masley's skepticism in a dispute over whether superintelligent AI could soon pose serious danger to humanity:
- Masley's argument: given the pace of machine intelligence, three simple facts — humans have no plot armor, human intelligence transformed the world, and intelligence isn't magic — should lead one to conclude "truly insane things could happen relatively soon," independent of adoption speed.
- PauseAI UK's stance: it takes no outlandish beliefs to see near-term superintelligence risk; assuming we're obviously fine requires the odd view of the world.
A typical collision between risk-aware and skeptic camps in AI safety, argument-based on both sides without new evidence.
More from AGI Musings
- Dev predicts an acceptable-quality AI writing harness within 3 months — tokenbender · 2026-11-12
- Dean Ball: 'It will not be AI in government. It's going to be AI as governments' — fkasummer · 2026-11-12
- Dario Amodei: AI Captures the Essence of Cognition, Biology and Math Discoveries Now Possible — haider1 · 2026-11-12
- tszzl rejects "winner-take-all AI": if true, Anthropic wouldn't exist — tszzl · 2026-11-12
- Epoch: AI software progress cuts training compute needs several-fold per year — datagenproc · 2026-11-12
- How to adjudicate if an intelligence explosion 'happened': GDP under 5%? — teortaxesTex · 2026-11-12