Claude profiles a simulation build: 1.5 of 2.2s comes from one pure-Python Newton function
andrew_n_carr · x · 2026-10-22
Andrew Carr shares how Claude profiled a model build: 1.5s of the 2.2s build time came from the pure-Python Newton function buildedgenringedgecollisionfilter, which walks each edge's neighborhood. Claude is now writing a vectorized drop-in replacement, doing what Carr says was 90% of his grad school work.
More from coding & agent
- Nous Research ships official Hermes Agent plugin for Meshy AI 3D generation via MCP — NousResearch · 2026-10-22
- Support AI Regurgitating Stale Docs Is a Growing Failure Mode — Give It Real Platform View — joshwhiton · 2026-10-22
- CallBaba launches BabaBench, a benchmark testing AI agents on real patient advocacy tasks — mattturck · 2026-10-22
- 70%-95% failure rates: devs say AI agents still too unreliable for real workflows — Luke17man · 2026-10-22
- How do you actually test agents before production? Devs tackle failure-path testing — Last-Bluejay-4443 · 2026-10-22
- A Reddit long-read proposes the Risk and Judgement pattern for bounding agent autonomy — mostly_deterministic · 2026-10-22