Agent Reliability Erodes Review Discipline

San Francisco — Simon Willison, developer and AI blogger, no longer reviews AI-written production code. He argues that as agents grow reliable, the habit of reviewing output atrophies — the critical s

Hacker News