Herbie Bradley: 'superintelligence transforms the world' arguments are motte-and-bailey
sebkrier · x · 2026-11-21
Herbie Bradley argues that 'radically transform the world' superintelligence claims often use a motte-and-bailey: the defensible motte (things get crazy, disempowerment seems likely) vs. the much weaker bailey (AI overtakes humans at the frontier of science). He notes that 'AIs will be at least as good as humans at every intellectual task' is correct but doesn't decide between Boaz's vision and full disempowerment, and calls for serious writing making the case that 'humans not at the helm' is actually likely.
More from AGI Musings
- Skeptics call the Claude Riemann proof 'hasty': the problem was easier than thought — gleech · 2026-11-21
- If Anthropic cracked AI R&D with 2x speedup, why is it still tied with OpenAI? — adam_karvonen · 2026-11-21
- With RH fallen, author says EA/AI safety is over-anchored on forecasters — JacquesThibs · 2026-11-21
- Gary Marcus: I got 'endless shit' for 3 years warning of the AI bubble — now everyone claims they knew — GaryMarcus · 2026-11-21
- Kevin Roose on racing to document AI history before secretive labs erase it — every · 2026-11-21
- Veteran pentester: AI hacks faster and better than I ever could — defenses can't keep up — yuridiogenes · 2026-11-21