Agentic Code Review

This article discusses the impact of advanced AI coding agents on software engineering, arguing that the hard part has moved from writing code to deciding whether to trust it. It highlights that code review is now the most leveraged skill, as agents produce code faster than humans can read it. The piece offers nuanced advice for different contexts—solo developers vs. teams with legacy systems—and notes that AI tools themselves can help manage the review workload. It cites data showing AI quadruples code output but only delivers 10% more value, emphasizing the critical role of review. The author is optimistic about agentic engineering but stresses that effective review depends on the developer's situation.