RC RANDOM CHAOS

AI Coding Agents Boost Output But Code Review Bottlenecks Increase

· via Hacker News

Original source

Harvard researchers find AI coding agents shifted the bottleneck to code review

Hacker News →

A Harvard study analyzing 718 firms found that AI coding agents increased code production by 30% but did not improve measurable work completion. Code review times rose 49%, with more change requests and comments per pull request. The research used data from engineering analytics platforms and LinkedIn records, tracking firms adopting AI tools between 2021 and 2026. The study highlights that faster drafting overwhelmed fixed reviewer pools and may have reduced draft quality, creating bottlenecks despite AI review tool adoption. Legal and compliance professionals are advised to evaluate AI systems using metrics like revision rates and reviewer time, not just output. The paper, still under peer review, suggests AI’s impact on code quality and review processes requires further scrutiny.

Read the full article

Continue reading at Hacker News →

This is an AI-generated summary. Read the original for the full story.