A recent study by Harvard researchers Fiona Chen and James Stratton found that the use of AI coding agents led to a 30 percent increase in code production among engineers, but the time required for code review increased by 49 percent. The study analyzed data from 718 firms and concluded that the rise in code volume did not correlate with an increase in finished work. This suggests that while AI can enhance coding speed, it may also complicate the review process, raising concerns about overall productivity.
Key facts
- AI coding agents increased code production by 30 percent.
- The time for code review rose by 49 percent.
- The study analyzed data from 718 firms and 725,938 workers.
- Finished work did not measurably increase despite higher code volume.
- The paper is currently undergoing peer review.
A short AI summary in our own words. The reporting belongs to Hacker News — Front Page; we don't reproduce it.