Superintelligence may excel at verifiable domains while having terrible taste elsewhere
dhadfieldmenell · x · 2026-11-21
A forwarded observation: people haven't fully thought through the implications of superintelligences being superhuman at verifiable domains while having very poor taste on non-verifiable ones — an asymmetry with real consequences for capability assessment and society.
More from AGI Musings
- Christiano: Two Months of Math Progress in a Decade, and ML Could Be Next — OwainEvans_UK · 2026-11-21
- AI safety researcher: doubting China's moratorium compliance is like doubting your own marriage — DavidSKrueger · 2026-11-21
- Two Riemann hypothesis proofs announced same day: one with Claude, one "all human, no AI" — Dr_Atoosa · 2026-11-21
- Hinton may be softening on an AI moratorium after SF risk-scenarios panel — DavidSKrueger · 2026-11-21
- Thinking Machines releases Inkling 2 Mini, a looped transformer topping sub-100B open-weights coding benchmarks — SonglinYang4 · 2026-11-21
- Mathematicians will study Lean proofs like N-body simulation results, researcher predicts — mjuric · 2026-11-21