Skip to content
assaynotes
EN日本語中文

Study: AI coding tools do not raise software output

Past event1 reports1 sourcesupdated 7 d ago

What happened

AI digest

Harvard researchers Fiona Chen and James Stratton analysed Jellyfish's engineering team analytics data, covering more than 700,000 employees at over 700 software development companies and 300 million work events (such as commits and pull requests) between 2021 and March 2026. The study concludes that AI coding agents generate more code but do not produce more software output.

Written by AI from the reports · updated 6 d ago

LatestOct 10 03:43
Based on data from over 700 companies and 300 million work events, Harvard researchers conclude that AI coding agents generate more code but do not raise software output.

Timeline

Follow the reports to see the event from each side.

Oct 10
  1. Ars Technica · AI
    AI coding agents generate more code, but not more software

    Study finds coding efficiency gains get "absorbed" by human review "bottleneck."

Heat of this event

Current heat 3·Peak in the comparable range 9（Oct 10 07:00）·Change over 24 hours in the comparable range -49%

02.557.510Oct 1007:00Oct 1022:00Oct 1112:00Oct 1203:00

The trend compares only the same accounts observed without a break, so its range can be narrower than the current heat count. Move the pointer or tap the chart to see each hour; with a keyboard, use the left and right arrow keys.