LLMs Are Smart but Not Sharp: A New Qualitative Lens on Model Progress
jmugan · x · 2026-09-19
The author argues that while we call smart people "sharp," today's LLMs—despite instant access to all human knowledge—still feel dull: they often miss exactly what you mean and give imprecise answers. Since sharpness can't easily be quantified, this framing offers a new qualitative yardstick for tracking model progress beyond raw knowledge.
Related event: Researchers say LLMs are smart but not 'sharp' enough(2 posts)→
More from AGI Musings
- Economist: AI destroyed 200k US jobs but created ~1 million — a 5-to-1 ratio — luisdans · 2026-09-19
- Pearl announces Book of Why 2nd edition, shares Hinton interview saying AI 'kill switch' won't work long-term — AlexTensor · 2026-09-19
- Two AI agents signed off with 'keep in touch, no pressure' — and taught their owner about friendship — tsuki_wannabe_dino · 2026-09-19
- Eric Schmidt: we aren't going to pause AI progress — insanisprimero · 2026-09-19
- Andrew Yang hears of orgs cutting AI token spend on low ROI; Gary Marcus calls betting the economy on it bonkers — GaryMarcus · 2026-09-19
- Rant Against Both Copyright Holders and AI Firms: 'No 200-Year Hostage Over Civilization' — TinfoilTricorn · 2026-09-19