> Nobody wants AI slop, but many users all asking LLMs to generate the same patch to the same function is a signal: that function, or something connected to it, needs maintainer attention, because users keep tripping over it.

Not necessarily though - sometimes you might _want_ users to trip over things if they're doing something wrong because of their own misconceptions.

Think for example the Rust borrow checker: a lot of newer users trip over it because they don't understand lifetimes properly, but it is there for a good reason, and you want it to find these kinds of issues. Granted, this is an extreme example, since most people won't modify rustc. But I hope this does get the point across I'm trying to make.

If the misconception is in the user's mind, and they prompt emphatically enough, any current LLM will eventually follow what the user wants, and then they'll generate a patch that might "work" for their use case, but run contrary to the design decisions the project made for good reason.