From HN comments:

Is it reasonable to imagine that LLMs should be able to play chess? I feel like we’re expending a whole lot of effort trying to distort a screwdriver until it looks like a spanner, and then wondering why it won’t grip bolts very well. // Why should a language model be good at chess or similar numerical/analytical tasks? // In what way does language resemble chess?

Logic + ability to correctly interpret the input.

Tasks like maths, chess etc., are good gauges for logical reasoning; and logical reasoning is essential for LMs to perform as language models - because it’s the difference between true statements and grammatically correct + semantically meaningless slop. As such, the LLM doesn’t need to play chess specially well, but it should be able to provide you not-blatantly-stupid moves.

It should be also understand the position correctly, based on Forsyth-Edwards notation; that’s linguistic ability at its finest.

Among all four models that I tested, only GPT provided a somewhat satisfactory answer (once you disregard that promoting the pawn to queen means to be checkmated in the next move; I’ve seen potatoes and HN users with more depth of reasoning).

I asked this one in the LLMs provided by DDG. The only one to correctly “interpret” the board was GPT-4o mini; Claude and Llama proposed the wrong moves, and Mixtral behaved as if whites were to play. GPT suggested to promote the pawn to a queen. ¬.¬

Lichess board analysis was however quick to point out the right move:

Yes. Even if I’m mildly interested on the underlying tech, I can’t help but roll my eyes and say “this shit again…” every time that it gets mentioned there.

For me that’s just an additional malus. My overall interest in HN has been dropping over the years; it reached a point that I deleted my years old HN account, and I mostly talk here about stuff shared there.

Specially my interest in the comments. And as a user there highlighted, “That is sad that you come for the posts and not the comments.” …yes, HN got sad. But then, look at the other comments and you’ll get why:

• Everyone and their dog repeats the same answers. No fucks given if adding to the discussion or not.
• Always got to have at least one assumer there sounding like “I assume you’re a hater”. Here it’s the “Are you boycotting LLMs or something? […]”
• All those answers that boil down to polite ways to reword “suck it up pumpkin lol lmao.”

That is not intellectually stimulating. It is not something that makes you to want to hack a subject apart, and know every single bit of it. And the fact that HN isn’t hacker-friendly any more is basically an open secret among the moderators there, that make absolutely no effort to change the situation, and instead would rather go after people highlighting the problems, “shoot the messenger” style.

Automatic ban? PTB unless 1) the ban is because of a domain associated with spam, and 2) there’s an appeal procedure for false positives.

If I’m reading it right, I also disagree with it, on a practical level; it’s like herding cats and not enough. I also think that an instance for women is the way to go, at least at the start.

However, to be fair with drag those actions aren’t mutually exclusive, and even if only a few admin teams keep a closer watch on sexism, it’s already improving the situation.

Likely for succinctness. I’ve also seen people listing only the nominative, specially when they’re OK with 2+ pronouns; e.g. “she/they”.

Feel free to use it for the Matrix community if you want! Or this sub. Or… dunno, any space that you want, to your hearts’ contents :3

Nah.

I’m saying that they should be blamed for condoning Nazism; not that they are themselves Nazi. The difference is important here because one does not automatically mean the other.

Also, the ones that are validating people to join the alt right are the people already in the alt right. Not the ones saying “you muppet you’re helping those fuckers by staying in this shithole”.

Lenny, you got the same problem as I do: writing huge walls of text, with really long sentences and paragraphs, that are really hard to parse. People lose track of what you’re talking about, so simply paraphrasing excerpts won’t help; you probably need to restructure the whole thing, if you want to be understood.

To make it worse, you’re redirecting the reader over and over, both through links and through indirect references.

This is on-topic here because I don’t think that db0 was power-tripping, but I don’t think that you deserved it. I think that it was just miscommunication; under a quick glance your post linking to buzzly.art doesn’t seem to be about rule enforcement, it sounds more like a rant plus request for advice.

And then [he] skips [them] altogether, because [they] are all contextually obvious.

I understood “the admins” as “a meaningfully large amount of the admin teams of the Fediverse”. Collective action.

I think that it would be a good start. But only a start; sexism is a social problem, so even if you ban the individuals saying sexist stuff, you still see sexism elsewhere.

And even if you ban overtly sexist users, others will keep:

• focusing on topics typically enjoyed by men, and typically disliked by women;
• interpreting what each other says based on masculine social norms;
• assuming that they’re dealing with other men unless explicitly told otherwise;

etc.

That’s still aggravating, you know? You can’t pinpoint why but it still makes you feel unwelcome.

Yeah, this sort of combo doesn’t work well. Passion fruit is also extra sour and it doesn’t mix well with citrus fruits. (Then you get guava crying in a corner, “I could’ve become guava cheese and spend the rest of my days over a slice of milk cheese, but nooooo, I had to try to be different!”)

Downvotes don’t mean “I don’t think this adds to the conversation” or “This is appropriate”, they mean “I personally don’t like this” here, and I think that kills a lot of our smaller communities.

Yet another nasty redditism inherited by Lemmy… and frankly that’s why I think that we should have multiple types of downvote, this way people can express their disagreement in a fast and pseudo-anonymous way without fucking everything up.

Yup, that’s a problem. Specially because, once the gender ratio gets too skewed towards one side (it is), the Petrie multiplier kicks in; then the sexism targets each woman more and more frequently.

Potential solutions that I see for the problem:

• Perhaps creating a few instances for women? I don’t mean instances to talk only about feminism, but for general stuff. With higher standards against harassment.
• Better mod policing against harassment. Collective action, so it’s easy to say and hard to do it, I know.

Start associating the site with Nazism, and the people still using the site as condoning Nazism.

PTB.

What you’re saying is accurate and relevant; you are clearly not trolling, contrariwise to the claim in the modlog. Your tone isn’t even impolite, it’s simply dry. And it’s geared towards the article, not towards other users.

All those “explanations” babbling about Musk’s “intentions”… aaah, when will they cut off the crap???

Like. Let’s pretend for a moment that Musk was totally trying to convey “my heart goes out to you” or crap like that. It doesn’t bloody matter - regardless of his “intentions”, the gesture conveys support to Nazism, and it’s the only sensible way to interpret it within that context.

Yup, something like this - but for the honeypot, not for the legit pages.

I should have clarified that I was focusing on an individual level.

On a collective level it’s exactly like you said, we [people in general] need to organise ourselves and get rid of this cancer.