AI can pass every test and still fail in production, the post argues

JnBrymn · x · 2026-07-21

The post argues that AI can generate code that passes tests yet still fail in production at the worst possible moment.

The core issue, it says, is not the code itself but that AI does not know what questions to ask about infrastructure, scale, and cross-cutting concerns—so the architect is still needed.

Original post →

More from coding & agent

coding & agent channel →