The OP put those addresses on that web page, and only on that web page. Some addresses received spam.

Edit: that’s not to deny that big data leaks are a serious problem

Clearly Chuck Peddle is a major designer and engineer. So what went wrong with the new computer he’s describing?

The company filed for bankruptcy protection within a couple years. https://en.wikipedia.org/wiki/Sirius_Systems_Technology

Steamrollered by PC compatibles obviously. At the time it wasn't clear yet that for 8086/8 you needed register level hardware compatibility, not just BIOS call compatibility (as in the CP/M days) to stay in the market. And nonstandard disk format to boot.

The non-standard floppy format was a huge annoyance for users. While the higher density formats were cool, the hardware could operate on PC-compatible format, but the OS wouldn’t support it.

ROM BIOS compatibility would have been nice, but it could be implemented at the custom MS-DOS version and run from RAM, but I’m not sure there were clean room implementations back at that point.

>If intelligence is necessarily coupled to a desire for self-preservation and self-interest, at what level of machine intelligence do the machines simply refuse to design their own more intelligent replacements,

At a higher level of intelligence than many humans, current experience suggests

If they take too much then confidence in the coin is absolutely lost and the coin fails and it’s price rapidly goes towards zero, so they’re possibly being smart by only taking a small percentage — if that was the hackers decision

Yeah $25m is only little but could still be useful

Maybe a player can learn which dice are biased then choose those dice to throw depending on what result would be best for them at that moment? So they gain a slight edge.

No, you don't choose dice to throw. The dice are all thrown and then you choose.

Rationally, you apply fines as close to the source as possible. Because they will pass those costs up the stack.

But the source could be the most likely place for corrupt reporting. Or: Maybe the source element is not dangerous but downstream by-products are.

Like you’ve said: It’s a problem.

Try asking for jokes about, eg Kenyans, Ugandans, South Africans

I think it might still refuse, but in your original test, German usually means a nationality, but African doesn’t.

I’m sure the jokes were terrible anyways

>AI-generated replies really are the scourge of Twitter these days

This is a complex problem. But the first step of that problem is Twitter/X

Avoid it, and the next step toward a solution may be easier.

HN is getting filled with AI generated articles and comments too. There's very few places safe from the avalanche of slop coming.

Yes! You are absolutely correct! (Pun intended)

That's true but it's just like with ICOs, the so-called Web3.0 and so on - there is a percentage of people aggressively promoting these, with a part of the community getting fascinated like with everything new, then with time novelty fades and people have a more balanced view of the new tech and these things get downvoted quickly.

I tend to agree, but seeing how the AI cult has turned into a veritable religion I struggle to share your optimism.

The solution is a social one. Most of the reason it's a problem in the first place is people defending/propagating slop as if it's worth something. The quantity isn't so high that community moderation can't handle it if it becomes socially unacceptable.

One way to combat this would be to force users to stake something. Pay 10 bucks to your account and if you misbehave by spamming or posting only AI slop, you lose it. Brings with it other problems, of course.

There's stacker.news - of course centred on Bitcoin discussion - that works on this principle. Posting and upvoting are actual microtransactions.

That's a nonsense idea because it fails to define how low-quality undeclared slop (LQUS) can accurately even be classified. Also, if money is on the line, it will be taken away even when the article is not LQUS.

I agree, but there is a slight alteration of the proposal which could work rather well. Pay $10 to get in, but no change to the procedures by which your account is revoked. This puts a price on sock puppets, while almost any legitimate, normal user only wants one account, and gets it for a trivial fee. This may also relax the pressure to monetize through ads, which could have perks.

Yep! This pretty much!

In fairness, the bigger problem as I see in comments is accusations of slop with zero evidence, often in an unfair attempt to suppress the takeaway message of an article.

Look at it from the other side: if Twitter/X gets swamped in AI slop, maybe that could be the end of it.

It's frying quite a lot of brains on the way down, sadly.

Also true! ;-)

Yes. I quit over a year ago. I don't miss it. It's a useless and toxic platform.