I think it is because it can't actually structure an argument properly, because it has no logic. It can't tell what is and isn't true, and so tries to simulate how ideas get conveyed, but without any ability to actually assess an idea and provide evidence for it or against it.

Turns out that learning the underlying meaning of sentences from a long stream of unfiltered internet is too hard to do by just wiggling a bunch of multiplication matrices. But you can get long chunks of accurate text predictions if you focus on every cliche that avoids communicating any meaning.

LLMs constantly do the rule of threes because it probably picked it up from motivational speakers or some shit.

The trainers probably think "muh nuance" was good and kept feeding it cookies.

I always see the comment “when will people realize it’s not left vs right it’s up vs down” under political tiktok videos and I want to strangle whoever wrote the comment. Now I’m sure it’s probably some bot shit

I'm assuming it's a product of the reinforcement learning; it must appeal to the people who rate responses and as a result got artificially dialed up from its ordinary background level.

I thought you were calling me out at first https://hexbear.net/comment/6911023

🫣

It's probably a human telling the AI agent to write a pissy blog post. But who knows.

They've deleted the blog post since (it's still in the git history), and the AI has written two more blog posts ... continuing the trend of wasting human attention even if more apologetic this time. That's the worst part to me, the person behind this is showing no sign of feeling responsible for the harm they causing with their AI agent. The apology is written in the voice of the AI agent. Nobody expects the agent to be accountable to that's worth less than nothing.

At this point people will have to prove they're human in order to even get a foot in the door (which possibly involves deanonymizing).

What if that's what pushing AI on everyone was about all along

Appearantly curl closing the bug bounty did very little to discourage slop submissions.

Dehumanizing people, wow really funny (this is sarcastic)

ⓘ This user is suspected of being a cat. Please report any suspicious behavior.

Talking to the AI as if it is a person is so fucking weird. But even weirder than that, is what they are saying. If for some reason you imagine that the chatbot is actually a "person" that you should address as a person, then imagine how fucked up it is to then explain that this person should refrain from commenting on policies about their own personhood.

Like, how do you convince yourself AI is actually people (it isn't), but then say that those people should stay out of conversations about their personhood? It's such a convoluted nightmare of brainworms that really shows how intertwined "believing a chatbot is a person" and "believing some people are lesser than others" are.

I hope we get to the point where maintainers don't feel the need to explain themselves why they close slop submissions.

But you kind of have to be courteous all the time if you want to foster a pleasant community. If you come across as irritable you attract trolls and harassment.

I don't even know how they best move on from this. I'm not familiar with the mpl code base but the changes are probably correct. Probably just merge an equivalent fix and ignore the undeserved hostility. Waiting for a newby to make the fix and get ridiculed by trolls just seems unnecessarily cruel.

I'll lock the thread, because discussing appropriateness of past behavior does not move the project forward.

Honestly the matplotlib people speak like robots anyway, full of manager speak.

python being the home for AI is the cherry on the shit cake. Every time I'm forced to interact with python my mental faculties are degraded beyond repair

This isn't even about pro or anti AI. This is about AI users actively sabotaging community building and outreach efforts and then slandering the core maintainers on top of it all.

I'm not even sure if the AI users just don't get it or whether they are actively trying to destroy open source communities. Or whether writing pissy blog posts is just part of the automatic routines.