Code Review in the Age of LLMs

A growing share of pull requests are now drafted by models. The bottleneck has moved: writing code is cheap, understanding it is not.

Review for intent, not syntax

Linters and models catch style issues. Human reviewers should ask: does this change do the right thing, and is it the simplest way to do it?

Smaller changes, always

Large AI-generated diffs are tempting. Break them up anyway. A reviewer who skims is not reviewing.

This is a sample article — replace it with your own writing.