Get the latest tech news
Harvard researchers find AI coding agents shifted the bottleneck to code review
Harvard study: AI coding agents raised code output 30 percent; finished work did not measurably rise, and getting code through review took 49 percent longer.
None
Or read this on Hacker News