Aiden Bai: human code review is the bottleneck — shift to verification loops
aidenybai · x · 2026-11-22
In a discussion with Sentry CEO David Cramer, Aiden Bai notes teams are already doing quality work with models (via opencode and others), typically with senior engineers setting up frameworks and systems first.
He argues 90% of human code review today is performative and humans are fundamentally a bottleneck for model-generated code. If frontier tokens got 100x cheaper tomorrow, he says you'd want to move humans to far higher-leverage work like building verification loops.
More from coding & agent
- Claude Code Code Mode: orchestrating multi-step tool calls via JavaScript for speed and context efficiency — AIFlow_ML · 2026-11-22
- Claude Code's new Code Mode runs multiple tool calls in parallel via JS scripts — AIFlow_ML · 2026-11-22
- 9 tweaks to make Google's Antigravity CLI actually usable with Gemini 4 — dotey · 2026-11-22
- Sentry CEO: cheaper tokens won't fix code review, vague intent is the problem — zeeg · 2026-11-22
- Adding WebMCP to a Static Hugo Site: One Tool Call Saves ~99K Tokens vs HTML — DanWahlin · 2026-11-22
- Sentry CEO: trust agents by code location — Sentry changes get full human review — zeeg · 2026-11-22