This! If you try to review agentic generated code in the way you review human developer you’ll waste time. LLM’s don’t learn like humans. So much of human code review is about teaching devs and developing a common team set of cultural practices.
RE: 
Bluesky Social
Gergely Orosz (@gergely.pragmaticengineer.com)
Unpopular opinion:
Current code review tools just don’t make much sense for AI-generated code
When reviewing code I really want to know:
- The...

