Story · Daksh Gupta (Greptile)
AI-Generated Code Is Already Competing With Human Code (Daksh Gupta (Greptile))
talk · Story page
Greptile compares agent-written and human-written pull requests on revert rates, flagged issue severity and review rounds.
In plain words
- Greptile found that software changes written by artificial intelligence performed similarly to human-written changes on the checks it studied.
- It checked how often changes were undone, how serious reported problems were, and how many reviews each change needed.
- Mistakes varied by tool, from security weaknesses to making too many separate requests for stored information.
- Software reviewers can use these findings to look for the kinds of mistakes each tool is more likely to make.
Appeared in
- Agent-written pull requests match human ones on revert rates, and fail differently
Sep 28, 2026 · lead story
Subscribe
Get the brief in your inbox
Pick daily, weekly, or both. Nothing is gated either way: every issue is on the site and in the feeds.
- Weekdays at 8:45am IST, one lead story and 6 to 9 items.
- Sundays, an argued synthesis rather than a recap.
- One click to leave, and quiet days say so in the subject line.