This is a question for code reviews inside a company, not public repos.

AI has clearly increased the throughput of code changes, which is good. Now you don’t just work on your main task but can also fix or improve small things on the side.

Let’s say before AI a dev would send 1 PR/day, now with AI it’s 2 PR/day.

The problem is that the review throughout did not increase. AI reviewing code is helping to remove obvious errors but you still need a human to do the final review, and that’s a bottleneck.

How’s everyone dealing with this problem?