
Code Review Is Overloaded

Stop being skeptical about AI for development with Charity Majors
Topic 6 of 17
Published Aug 12, 2026 · Clip 22:14–26:56 (4:42)
Charity explains that code review is overloaded—it bundles mentoring, product decisions, and bug-finding—and argues validation is at the bottom of what code review should be. Good parts: deciding 'do we want this in our product,' API design, and mental-model discussions; weak parts: reading for syntax and bugs. She praises Intercom as a high-watermark for pre-AI companies becoming AI-native: AI-validated PRs carry the wisdom of their most senior engineers, freeing humans to discuss direction instead of nitpicking.