swyx hails Opus 5.5 as best work-estimator model; dev counters that review time is the real bottleneck
petrusenko_max · x · 2026-10-23
swyx praised Opus 5.5 as the best work-estimator model he's used: as delegation grows, models shouldn't quote human timelines of 2-3 days but 2-3 hours matching their parallelizable capacity. Max Petrusenko pushed back with a key caveat: agents finishing fast is real, but what no model estimates is human review time — the diff waits for him to read it, and the 2-3 days come right back. He wants two numbers tracked: agent time and review time, and 'the second one is the ship date' — a sharp observation that human review is the delivery bottleneck in AI coding workflows.
More from coding & agent
- Drop a PowerShell Script in a Folder, Get a Live MCP Tool — dfinke · 2026-10-23
- Auth'd MCP servers can deliver user-personalized skills, an overlooked design win — dsp_ · 2026-10-23
- Universe badge pairs RP2350 with an AGENTS.md so Copilot can code and deploy to it — mariorod1 · 2026-10-23
- Claude Code banked reset vanishes hours early, user demands Anthropic restore it — burny_tech · 2026-10-23
- Sudo L7 benchmark uses 76-word natural prompts vs 291-427 words for rivals — echen · 2026-10-23
- Claude recolors an AI-generated SVG into 4 sneaker colorways without touching stitching or shading — socialwithaayan · 2026-10-23