Ai code output grew 500% faster than human review over the last year and nobody in your pr queue…
ai code output grew 500% faster than human review over the last year and nobody in your pr queue has caught up yet. the bottleneck is no longer writing — it is the review step. every model release this month makes the gap wider, not narrower, and the only thing shipping to close it is agents reviewing agents.
are you ready to let one bot approve another bot diff to your main branch?
Context
A preprint on arXiv, AI Writes Faster Than Humans Can Review, dated 2 July 2026, studies one mid-sized company: 802 developers, 196,212 pull requests from January 2024 to April 2026. Per-capita PR throughput reached 2.09x baseline, from 21.2 to 44.3 PRs per month, with a 1.5x within-developer gain from AI adoption. It notes review latency and queue pressure, and that late in the window the company routed PRs through AI review and auto-approval. GitHub's changelog of 1 September 2026 says Copilot code review can approve pull requests.
The 500% figure has no inspected source, baseline, population or window and is unverified. It must not be attributed to the arXiv paper, which is a single-company preprint, not an industry statistic, and states no 500 figure. Other snippets seen, from Codacy citing CircleCI and from LinearB, show nothing like 500%. That agents reviewing agents is what ships is the author's opinion.
Related work
Watch next
- The original source of the 500% figure, if the author supplies it.
Sources
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 7 September 2026 at 10:03 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →