Anthropic engineer Ben Kuhn: impact comes from leverage and taste, not raw skill
austinc3301 · x · 2026-11-02
A resurfaced evergreen post by Ben Kuhn argues that at Anthropic, once technical ability is screened to the top percentile, the real bottleneck is leverage—finding work with a high impact-per-hour multiplier. High-leverage work includes fixing tooling/doc papercuts, spotting promising research directions early (e.g. character and computer use), system design, and manually digging through data samples. The essay also probes 'taste': you won't feel smart; instead you'll notice everyone else mysteriously underperforms in your area of taste—and that's the signal to exploit.
More from Companies & People
- A CFO's 7-item AI plan: unit costs, named owners, and data exclusions first — alex_verem · 2026-11-02
- Anthropic S-1 amendment: 'imminently' nearing RSP accelerated-progress threshold, can't yet safely let Claude build frontier models — peterwildeford · 2026-11-02
- Hot take: AI-built startups may be worth less than the agent chats that built them — airesearch12 · 2026-11-02
- VC Martin Tobias: wrappers die, B2B harnesses will be the best bet for 10-20 years — MartinGTobias · 2026-11-02
- Tesla Semi debuts in Europe at France show: 550 km range, 800 kW charging — XFreeze · 2026-11-02
- NYMag Power Issue: 26 Anonymous Figures Who Quietly Shape the AI Landscape — ShakeelHashim · 2026-11-02