AI's Impact on Code and Review

Stacked diffs and tooling at Meta with Tomas Reimers

Topic 5 of 9

Published Apr 2, 2025 · Clip 39:50–52:59 (13:09)

Thomas argues AI will lead to more code being written, creating second-order effects: more review, testing, merging, and deployment. He predicts AI will handle mechanical review checks (like a Grammarly for code) while humans focus on intent alignment and business decisions, since companies won't want machines making business calls. Both host and guest worry leaders may undervalue human code review because its absence is invisible until things break, and they discuss testing's evolving role as AI agents write their own tests.