Code review works better when several AI models must agree
A small team testing on its own code found a weak spot in s. When one model writes code and then reviews the same change, it can miss its own mistake and approve bad code. The team changed the review step so several models check each change independently, from different angles, and the change only moves forward when they clearly agree.
The key rule is that a comment does not count as approval; only an explicit approval counts. If any reviewer is unsure, the change waits instead of being merged. In one case, the reviewers went through about 31 rounds before they reached agreement, and the final result was a clean two-line fix.
This is still only an internal experiment with no outside users, so it should be treated as an early practice note, not a proven standard.
Key points
- One model reviewing its own code can approve mistakes it created.
- Several models reviewed each change independently and had to clearly agree.
- Comments did not count as approval; only explicit approval counted.
- from any reviewer paused the change instead of letting it merge.
- One internal case took about 31 review rounds before a clean two-line fix was merged.