I have no unique perspective to add other than an obvious question: If the PR is low quality, why not just close/reject it? Does it matter if it's AI assisted or not?
Because PRs with AI need to be reviewed with a lot more scrutiny, simply because AI is good at generating code that looks good, but isn't necessarily correct.
So now you're looking at a PR that at face value looks good, but doesn't reflect the author's skill and understanding of the subject.
Meaning now you shift more work to the owners of the codebase, as they have to go through those verifications steps.
If a person writes 100 lines of code there’s a (valid) assumption that someone thinks this 100 lines of code is worth writing. With AI it takes no effort to write 10,000 loc. Asking someone else to figure out if that code is worth merging just offloads the effort to someone else who didn’t ask for it.
It makes sense, because humans and AI write code different. Because humans aren't AI. The types of mistakes AI writes into code would never be done by a human.
Everything that a maintainer would need to prove to themselves to merge it can be codified in a pipeline.
Or some kind of protocol for building those things in the MR so that any new behavior explicitly demonstrates the new states and transitions.
This is hard if the new MR introduces a completely different paradigm outside the mental model of the reviewer and maintainer. Might be better off completely forking the project and running it in parallel aka taking on the maintainer duties if they feel so inclined to completely change things
If AI writes a for loop the same way you would... Does it automatically mean the code is bad because you—or someone you approve of—didn't write it? What is the actual argument being made here? All code has trade offs, does AI make a bad cost/benefit analysis? Hell yeah it does. Do humans make the same mistake? I can tell you for certain they do, because at least half of my career was spent fixing those mistakes... Before there ever was an LLM in sight. So again... What's the argument here? AI can produce more code, so like more possibility for fuck up? Well, don't vibe code with "approve everything" like what are we even talking about? It's not the tool it's the users, and as with any tool theres going to be misuse, especially new and emerging ones lol
I don't know why you have to qualify your sentence with "think carefully before you respond" it makes it seem like you're setting up some rhetoric trap... But I'll assume it's in good faith? Anyway...
I don't mind if a review is AI-assisted. I've always been a fan of the whole "human in the loop" concept in general. Maybe the AI helps them catch something they'd normally miss or gloss over. Everyone tends to have different priorities when reviewing PRs, and it's not like humans don't have lapses in judgement either (I'm not trying to anthropomorphise AI, but you know what I mean).
My stance is same about writing code. I honestly don't mind if the code was written `ed` on a linux-powered toaster from 2005 with 32x32 screen, or if they wrote it using Claude Code 9000.
At the end of the day, the person who's submitting the code (or signing off a review) is responsible for their actions.
So in a round-about way, to answer your question: I think AI as part of the review is fine. As impressive as their output can be sometimes be, it can be both impressively good and impressively bad. So no, only relying on AI for review is not enough.
So now you're looking at a PR that at face value looks good, but doesn't reflect the author's skill and understanding of the subject.
Meaning now you shift more work to the owners of the codebase, as they have to go through those verifications steps.