[Asraelite]

In principle I don't see a problem with bots looking out for grammar mistakes and typos that could confuse readers (so long as it's less intrusive than this bot), but in this case the bot is just incorrect.

"Good to merge? Test suite passes locally?" is perfectly valid English. You need to make sure that the bot is configured to not insist on arbitrary prescriptivist style guides that nobody cares about.

[samrus]

The problem is these bots dont understand the concept of "goals" or "achieving something". So they cant judge for themselves when they are actually helping (like in correcting grammar where its so bad its hindering communications with other contributors) or being pedantic (like this) because they wont think to consider the difference unless wxplicitly told to do so

And NGI like a human would be given this task and consider the spirit of the law, the goal we want to acheive, and enforce it to reach that. The next token AI doesnt model that. It just predicts the next token, and understanding the spirit of the law does not seem to be in the emergent capabilities of that

[Asraelite]

I often see comments on Github issues where poor wording makes it difficult to understand what is actually meant. Things like "I reproduced the bug on Linux, then I tried Windows. I can't reproduce it now." Does that mean it's just not reproducible on Windows, or not reproducible at all anymore? Ambiguities like that are especially annoying when it's someone posting a solution to a problem. Sometimes it's because of grammatical errors, sometimes not.

I think LLMs are actually great for catching things like this, and you generally don't need some higher-level understanding about the goals involved to notice the ambiguity. My point wasn't that bots shouldn't be used like this, just that they need to be given the right instructions.

[antonvs]

Yeah, no-one wants Grammarly on pull requests.

[samrus]

Well the maintainers might. The both in (badly) enforcing guidelines they set out. So they might want to police grammar a little bit. And i could see their point if the grammar is so bad it hinders communication. Bit uncomfortably, it was also be a reasonable predictor of those spam PRs put in from people just trying to pad resumes.

But yeah the bot needs to loosen the boundary a bit there