Applejinx

But that's exactly it. It doesn't 'know'. It isn't. It's a possible story nobody told. There is nothing there.

I've never seen anybody or anything go out, suck up everything in the world and distill it into a different thing containing elements of all the things they took for no particular reason with no expectation or intent.

That's a WILD angle for them to argue. Lotta energy being expended for a lack of intent. Lotta money spent…

I typed the first stanza of 'Jabberwocky'.

Worked about as well as expected :D

…as far as you know, at the first glance.

Could you work out that it was wrong, given weeks to go and check its work? If so, your trust is misplaced.

You ARE taking days or weeks to go and check, yes?

VERY MUCH to be confused, which is exactly why Epic will doubtless be having many w… no, I'm sorry, I think they'll have one word. 'Nope'

Likewise. This seems to be the norm for AI people: act as if.

No, language is communication. Those numbers just sit there unless someone tries to get them to do something. There's nobody in there and nothing to communicate, and you might fairly ask whether, if that sand WANTED anything, it wished to just be let alone without being made to jump through hoops for the pleasure of humans.

Only at the first derivative. What you're observing is an additional layer of 'but, in genre X!' and then same slop, different prompt.

You're complaining about what happens when people don't even supply the additional layer, but just go 'make me a poster, no mistakes, it should be a great poster'. When you give no direction at all you get the purest of slop. You start giving directions, you start narrowing down the area you're digging through the slop of, and possibly you ask for something that people aren't familiar with and they're only able to identify that you asked.

Hence, Chinese posters with a cherry blossom and the kanji for 'Japan' and such. Someone went 'make it Chinese' and the slop associated it with 'Oriental' and what came out was collective unconscious, such that people who weren't thinking and didn't have experience might not look twice, but the meaning turned out tainted.

Sort of like food depictions that more closely resemble a pile of maggots. Those glisten! (well, I suppose they must.) There's a whole uncanny-valley thing that happens on every level and inspires revulsion. I'm suggesting that is still happening in the other examples but you've not got the experience to register their problems. With the food depictions, all animals eat stuff so we're hardwired to bounce off unsafe foods even if they glisten…

If there's nothing to think about they may well not think.

That's the problem.

Why would any LLM 'think' in terms of trying to cite prior work?

It itself is prior work. It's asking a fish to show where the water is. The fish can't imagine that absence, and the LLM can't imagine anything not being prior work.

OK, how about 'nothing is real, nobody will ever care again, let's all never listen to anybody else ever again'?

That's more or less the same thing expressed more nihilistically. But nihilism is, like, just a word man (to 'The Dude' it up a bit). Perhaps we are either fiddling on the deck of the Titanic, or just staring at the waves?

What do you think the results will be of the machines consuming only their own slop and Habsburging themselves intellectually, and then all the humans abandoning any attempt to either talk or listen? As a global civilization of humans? I'm not sure we've explored the likely consequences of 'if you can't spend them time to write it why should anyone read it'.

Maybe it's gone too far already.

When I saw the negative score I went 'Hmm! soft spot!'

Of course calling out something like that would get a response. Those who are interested in furthering 'the natural world for white humans' will immediately see the possibilities of this sort of motte-and-bailey stuff. Fairly often they're directly working that beat and are very familiar with what I'm suggesting: I wasn't saying it to bring it to their attention, they're quite aware.

I mention that the definition of 'human' matters, because there are a bunch of people who don't consider skin color a disqualifier. And I spoke up because I'm such a person, and I'm wary of ways to sneakily redefine such words until common usage includes such caveats.

I would like to have more clarity on what it considers 'human' and 'the natural world' because you could use that framing to run with a really wild ultra-right-wing viewpoint where only extremely white people are human, and the natural world means scientific medicine must be destroyed.

We don't know what it's up to unless we know how it defines these terms. What's 'primacy'? I would say climate has primacy over the artificial constructs of human civilization, 'cos we're able to nudge climate in some very alarming directions we're ill-suited to protect ourselves from.

From what I've seen of other AI-enabled projects interacting with forums in my field of work, it's very common for the AI people to jump in and answer every possible question and respond to every possible interaction, very quickly.

Could be just a bot doing it, but it could also be a sign of an inexperienced person trying to engage with an existing community without being part of that community. I know in my area it took years for me to learn that I shouldn't jump into every conversation, and that doing that rubbed people the wrong way, and I've seen others (many years before AI) wreck themselves by being unable to learn this lesson.

In some places responding bumps the thread so it's outright manipulative. HN doesn't sort threads like that.

I'm not sure this counts as astroturf, or simply evidence of a person who's not okay with letting a community opinion independently evolve. You could say 'well it's his job to fight the community if it looks like it's turning hostile' but think about that for a second. It takes some experience to know when you can let negativity be out there in the environment, and I think there'll be a lot of people who aren't capable of making such allowances.

I guess in ten years we'll see which is the out of touch Kung Fu master.

I do see a whole set of people doing mystical passes in the air, sometimes with wires holding 'em up, in a set of ritual beliefs propped up by vast accumulations of power, money and pride, not to be questioned.

Didn't expect Carmack to be among those waving his hands about.

If it's so amazing, let's see the amazing thing he does. Go ahead. I'm sure Doom needs a really on-point web page with Inter.

Yes, absolutely. Even human intelligence stops being useful at a certain point and becomes 'enough, provided the human is motivated and tenacious'. Intention is a better metric, but it's a serious AI weak point if not actively an achilles' heel. 'Warios' come to mind.

He could be lying in hopes of stalling everybody else.

You can say it, and there's no reason you shouldn't dare say it, but getting other people to believe it is a whole other story. Good luck with that. Guess we know what's in your wallet :)

Sure is for bass. It's a sealed box that you're in, with shapes that stop anything from acting like a simple room shape, and surfaces that can soak up highs, and so small that this absorption is a huge percentage of surface area.

There's a reason mix engineers have always done 'check the mix in a car'. It might be way more revealing of what's happening in the bass, even when the speakers are relatively weak. Once you get anything decent, watch out. It's a bit like having your whole body inside headphones.

Or, y'know: voting instructions.

The government does! At least right now in the United States. And yeah, I think it's reasonable for people to start asking.

The ones (perhaps not so much candidates but national leadership?) that outright pursue Russian interests are surely 100% corrupt, if you count 'semisecretly serving an entirely different and hostile country' as corrupt. I get that the counterargument is 'yes but USAID and NATO are existentially evil and so anything for the purposes of destroying NATO is legitimate', but at that point you're Sephiroth-posting and I don't feel obligated to respect that at all.

It's very easy to elicit this from LLMs. Anytime you've played with an LLM by typing weird stuff to freak it out, and got spooky results, it's that you've done. You've turned the story into a scary rogue computermonster story and that's all that has happened.

When these stories start to direct real-world activities, people in reality suffer, to even a catastrophic extent, and yet that's still all it is. Language models retell our stories, nothing more. And that is also quite enough to be worrying.

This is the same as systemic bullshitting. Not the first time I've smelled it on fluffy LLM output. I think it's the result of the training trying to induce the LLMs to talk over users' heads even when they are professionals, to entice further use on the grounds of 'oh it's so smart I can't understand its genius train of thought', but I don't think eliciting language like that really taps into 'associations of smarter previous language users'. More likely it's 'associations of rampant bullshitters'.

I do this! Interesting to see someone else advocating the practice. In my open source development I start work on a retro machine running Snow Leopard, specifically to be able to generate authentic retro binaries that'll work anywhere, and then I have a process to port to modern MacOS and signed Apple Silicon binaries, but the machine's a M1 Studio on Monterey and building for 11.1.

My worry is that they'll kill signing for people who aren't updating, though that shouldn't be a factor. I would have to simply get another machine to port to, because I don't trust that changes they make are always going to be good in any sense.

There might be a window for computing within which it's good: I find I'm not worried about adding new Linux machines and architectures to build for, but my Windows builds are 32 and 64 bit on Windows 7 with no attempt to modernize the build system, and I've got that two-stage Mac build process where it starts on a time capsule machine not connected to the internet, and continues on the M1 running Monterey. I'm anxious for the day Apple requires you use stuff that's constantly changed out from under you. I don't work like that.

Love the em-dash. Chef's-kiss.

A couple answers…

Why do you think for even a moment that they would buy only one copy, rather than every available copy of something rare and difficult to get? Anyone doing this stands to gain from destroying the only copies of something that has scarcity, to stop competitors from ever getting access to it. That's a serious concern.

Why would they do that? What is this 'should' to a company that would do this in the first place? Are you not suggesting the opposite of what they aim to produce?

Technically, killing everybody is one way of solving famine and poverty.

How weird that all those humans seem not to trust the logic of their betters ;P

I asked some AI-using compatriots a while back who were complaining about this, 'isn't it doubling down on bullshitting you?' and got some pushback along the lines of 'it isn't a person therefore doesn't have dark motives like that therefore can't be doing that to us'.

Didn't convince me. I think bullshitting like this can be a behavior, not just the intention of a human. If it's blowing a lot of smoke to use fancy words and phrasings (and semicolons! All the trimmings) it's fair to ask if it's systemically bullshitting you: i.e. the behavior is meant to have you shut up and trust it and not ask questions.

Who's driving that is still important: if the company's directing it to do that in system prompts that are adversarial to users, that's a big yikes. If it's an epiphenomenon of the company demanding it get ever smarter, maybe it's a sign that their demands are not having that result, rather they're making it bullshit more explicitly and mimic more 'smart' signifiers.

Wait.