swyx: low-effort plus critique loop beats max effort for throughput-bound agent work
alvelda · x · 2026-11-10
swyx argues that for throughput-constrained agent work, low effort plus a critique loop beats one-shot max-effort runs, noting that Karpathy also runs at medium effort. The practice is documented in his open-source swyxio/skills/babysit-runs. alvelda suggests an easy test: run your next batch job both ways and compare cost and quality.
More from coding & agent
- Agents build Lean models of every CPython 3.15.2 function, validated on 50,000+ generated inputs — ChrSzegedy · 2026-11-10
- 10 open-source repos that review your code with AI, from Alibaba to Anthropic — victor_explore · 2026-11-10
- The dialectics of AI coding: LLMs can't write prod code unsupervised — neither can humans — ShenRaphael · 2026-11-10
- SteerCamp's multi-agent crew holds daily standups, reporting progress and blockers like employees — pramodk73 · 2026-11-10
- Running Codex for two days to reimplement an outdated Android book's projects and retake screenshots — burkov · 2026-11-10
- Unverified: Fable 5.2 tops FrontierCode 2.0 at 49.1%, half the cost of Opus — FinanceYF5 · 2026-11-10