kelnos

SO's new-question rate peaked in 2014 and has been declining since 2017 (with a small, short-lived, early-COVID bump). LLMs have certainly hastened its decline, but it would still be in a pretty sorry state today even without them.

I'm pretty sure these days it's 95+% immigration, though.

> Why is it unreasonable to say, "I'd prefer something other than USian"?

Not the person you're replying to, but my take on it: because in the grand scheme of things it's small potatoes, and putting even a minor burden onto someone else to change their language seems unreasonable and unnecessary.

> The consequences will be me having a slightly lower opinion of you as a person.

That's an incredibly naive take. In many places, especially professional ones progressive/liberal/woke places, the consequences are much more than you having a slightly lower opinion of someone.

> I don't understand this compulsion to make either party into some villain.

If someone replies as the GP suggested, they will be villainized by some subset of the people who hear that reply. In a professional setting, that could involve a visit from HR and/or be a career-limiting move, depending on the company culture.

It doesn't matter if you understand it or not. It happens, and that's what matters.

> It's absolutely not the most demanding demand you could place on another person, but it isn't nothing either

Another way to think about this is as a parallel to microaggressions. Little things that may seem innocuous or not that big a deal on their own, but that add a little bit of emotional or cognitive load to someone's day.

"It's a good change!"

No, STOP is a keyword that the carrier intercepts. The recipient never sees it. The carrier will ensure that you don't get any messages on that number anymore. The business behind the number can (sometimes, depending on the platform they're using) set a custom reply to STOP, which is what makes it seem like they're getting the message, but they aren't.

Of course, it only applies to that phone number, so they can still spam you from another number in their rotation.

"Dead" is a relative measure. When I used it heavily (roughly 2004-2015 or so), most of my friends were active on there, and would post things (interesting or otherwise) multiple times per day or at least near-daily.

Nowadays I sign into my FB account every 3-6 months or so to find something specific about someone, and I see that my feed is dominated by 4-5 people (out of around 1,000 "friends"). If I'm curious and start searching for other friends specifically, nearly all of of them have not posted in years.

Sure, there are some active community-type spaces there, but that is a very small part of what FB used to be.

Instagram is a garbage platform now, too. I used it back when it was just a feed of the people I follow. Every few months I'll click an Instagram link sent to me by someone else, and then I'll check my feed out of morbid curiosity, and it's 70% promoted/recommended posts from people I don't know.

Glad I don't spend time there anymore either.

The only true defense is actual open source. Open "weights" is not that. You need the entire training pipeline open-sourced, plus the entire training corpus, so someone can independently train the model from scratch.

Obviously this will never happen. And even if it did, it still probably wouldn't be that hard to hide something in that mountain of data.

> But I trust a random human enthusiast and community feedback more than I trust a megacorporations that commit crimes without any prosecution or consequences.

I'm not saying we should trust megacorps, but they have a very strong incentive not to build these sorts of backdoors into their models; if that backdoor were discovered, their reputation would be shot, permanently, and they'd stand to lose out big, much much much more than any possible benefit they'd derive.

"Random" is the operative word in "random human enthusiast". I don't trust randoms on the internet, period. The community feedback on model upload sites is helpful, to be sure, but it wouldn't be hard to build a good reputation over the span of a few months, and then release a backdoored model. Given the pace of things here, that feels significantly easier than, e.g. getting a malicious commit into a popular open source project.

It sounds like you've defined "slop" as "anything an LLM produces". Seems unlikely you're going to get much agreement here.

I don't think that solves it. In your "example A", the compromised software can display that you voted for candidate A, but when doing the signing, sign the result that you voted for candidate B. There's no reason why the thing that it signs has to be the same thing that you see on the screen, and there's no way for you to verify it.

Perhaps, but why were the owners of the account then unable to sign into it after it was restored?