Conworld random thread

[WeepingElf]

Indeed! It is the techbros and the autocrats that are the problem, not the technologies themselves.

There we will simply have to disagree. There are some technologies that I would consider emphatically dangerous in themselves with no legitimate uses. Consider something like nerve gas. You yourself have criticized nuclear power so presumably you would agree at least some of the time.

Sure, there are harmful technologies that have no place in a good, sustainable future, such as bioweapons or nuclear power. Or, for that matter, fossil fuels. Such technologies have to be banned.

[jal]

Indeed! It is the techbros and the autocrats that are the problem, not the technologies themselves.

There we will simply have to disagree. There are some technologies that I would consider emphatically dangerous in themselves with no legitimate uses. Consider something like nerve gas. You yourself have criticized nuclear power so presumably you would agree at least some of the time.

I took this as "the technologies" specially referring to the aformentioned AI and biotech, not all technologies.


JAL

[WeepingElf]

Indeed! It is the techbros and the autocrats that are the problem, not the technologies themselves.

There we will simply have to disagree. There are some technologies that I would consider emphatically dangerous in themselves with no legitimate uses. Consider something like nerve gas. You yourself have criticized nuclear power so presumably you would agree at least some of the time.

I took this as "the technologies" specially referring to the aformentioned AI and biotech, not all technologies.

And that's precisely what I was intending to mean.

[Raphael]

Bumping this because it is, now, the last post on an earlier page of this thread, and people might therefore overlook it:

Interesting blog post by Bret Devereaux on taverns where adventurers meet, and how plausible or implausible they'd be in real life:

https://acoup.blog/2026/09/04/fireside- ... er-4-2026/

[malloc]

My point was that I don't want this setting dominated by AI despite taking place in the future. Thus it needs some explanations for why the current onslaught of AI ultimately failed and even reversed. Putting aside all questions of whether AI is good or deserves to take over, I just don't consider it my vision for the setting.

[WeepingElf]

My point was that I don't want this setting dominated by AI despite taking place in the future. Thus it needs some explanations for why the current onslaught of AI ultimately failed and even reversed. Putting aside all questions of whether AI is good or deserves to take over, I just don't consider it my vision for the setting.

Nobody knows whether AIs will ever start making money instead of burning it as they currently do. Currently, there is a big AI hype because everybody expects them to become profitable at some point in the near future, but they aren't profitable now. That is a speculative bubble, and such bubbles tend to burst - happened often enough in the past. And they are already facing opposition; you are not alone in worrying about them, and many of the people worrying about them try to do something about it. People don't want AI slop, and politicians (except of course the far right) call for regulating them. As has been said multiple times in various threads here, the abilities of AIs are currently vastly overrated. There is no artificial general intelligence, only algorithms that excel at some tightly circumscribed task, such as playing chess or predicting the folding of proteins. They are like savants - people who, for instance, can multiply six-digit numbers in their head but not tie their own shoes. Did savants take over the world? No. Will artificial savants take over the world? Probably not, either. We will most likely get a future where AIs exist and are used for a number of tasks, especially "bullshit jobs" such as writing the kind of business reports that are required by law but are just filed away and nobody will ever read unless something goes wrong, but do not rule the world. Read Bruce Sterling's story The Beautiful and the Sublime which portraits a future where AIs have freed humans from the burden of such jobs such that they are free to pursue their creative interests.

[jal]

Bumping this because it is, now, the last post on an earlier page of this thread, and people might therefore overlook it:

It was an interesting read.

There is no artificial general intelligence, only algorithms that excel at some tightly circumscribed task, such as playing chess or predicting the folding of proteins. They are like savants - people who, for instance, can multiply six-digit numbers in their head but not tie their own shoes.

I disagree, although I'm equally pessimistic about AI as you are. There is indeed no AGI, and there won't be in the forseeable future, simply because LLMs are Chinese rooms. However, they're not savants either, they're not just playing chess, or just folding proteins. They have mastered the human language, and are able to respond to it like a human would (sort of), beating the Turing test (or at least initally, nowadays it's pretty easy to spot an LLM text or reply). There's no intelligence at all in the answers, because an LLM doesn't understand what they're saying, and LLMs can't hold long conversations or ruminate over what's said before. But they are able to have what one could call a conversation, and act seemingly intelligently, something a savant would be hard-pressed to do.


JAL

[WeepingElf]

Bumping this because it is, now, the last post on an earlier page of this thread, and people might therefore overlook it:

It was an interesting read.

There is no artificial general intelligence, only algorithms that excel at some tightly circumscribed task, such as playing chess or predicting the folding of proteins. They are like savants - people who, for instance, can multiply six-digit numbers in their head but not tie their own shoes.

I disagree, although I'm equally pessimistic about AI as you are. There is indeed no AGI, and there won't be in the forseeable future, simply because LLMs are Chinese rooms. However, they're not savants either, they're not just playing chess, or just folding proteins. They have mastered the human language, and are able to respond to it like a human would (sort of), beating the Turing test (or at least initally, nowadays it's pretty easy to spot an LLM text or reply). There's no intelligence at all in the answers, because an LLM doesn't understand what they're saying, and LLMs can't hold long conversations or ruminate over what's said before. But they are able to have what one could call a conversation, and act seemingly intelligently, something a savant would be hard-pressed to do.

Fair. The comparison of AIs to savants may be not very apt, but I still think LLMs are overrated, even if there are reasons to worry about them (no, they won't take over the world; but yes, they may cost many jobs, yes, the servers gobble up a lot of power, and yes, they produce lots of shoddy artwork), and they ought to be tightly regulated.

[jal]

Fair. The comparison of AIs to savants may be not very apt, but I still think LLMs are overrated, even if there are reasons to worry about them (no, they won't take over the world; but yes, they may cost many jobs, yes, the servers gobble up a lot of power, and yes, they produce lots of shoddy artwork), and they ought to be tightly regulated.

100% agree. (Though they could "take over" the world for a certain definition of "take over", at least wreaking or taking control of the digital infra structure.)


JAL

[WeepingElf]

Fair. The comparison of AIs to savants may be not very apt, but I still think LLMs are overrated, even if there are reasons to worry about them (no, they won't take over the world; but yes, they may cost many jobs, yes, the servers gobble up a lot of power, and yes, they produce lots of shoddy artwork), and they ought to be tightly regulated.

100% agree. (Though they could "take over" the world for a certain definition of "take over", at least wreaking or taking control of the digital infra structure.)

Indeed, I expect the end of the Internet as we know it within the next 10 years. More and more web sites break down under the onslaught of the AI scrapers, and the techbros seem determined to replace the Internet by a "Metaverse", which apparently will follow the more profitable model of the centralized, hierarchically structured online service like AOL or CompuServe (and filled with AI-generated content, of course). And authoritarian governments will welcome such a move because a centralized online service would be far easier to monitor and censor than the rhizome-like Internet. However, in free countries, there will probably be a counter-movement to this with the goal of establishing a democratic, non-commercial alternative to the Metaverse free from AI slop. (ObConworld: This scenario is part of my history of the near future.)

[Raphael]

I disagree, although I'm equally pessimistic about AI as you are. There is indeed no AGI, and there won't be in the forseeable future, simply because LLMs are Chinese rooms.

The problem with the Chinese Room argument is that it assumes as a matter of course that, if you have a system that consists of a number of individual components, and none of these components is intelligent in itself, then somehow, this automatically means that the system as a whole isn't intelligent, either. Which is a problem because, well, individual human neurons aren't intelligent, either: https://zompist.com/searle.html

In any case, if, hypothetically speaking, you've got something that can work on all the practical tasks an actual general intelligence can work on, than I don't see how it changes much if someone can come up with a really clever philosophical argument for why the thing in question, on a deeper philosophical level, doesn't "truly" understand anything. It can still work on all the practical tasks an actual general intelligence can work on.

[Ares Land]

My point was that I don't want this setting dominated by AI despite taking place in the future. Thus it needs some explanations for why the current onslaught of AI ultimately failed and even reversed. Putting aside all questions of whether AI is good or deserves to take over, I just don't consider it my vision for the setting.

That's a very good point.

As this is conworlding, the question isn't really about predicting the future of AI, but about whether you can have a future setting without AI in a way that adequately suspends disbelief.

I think such constraints make for better conworlding (I think Dune is all the more vivid for the Butlerian Jihad.)
Imagining a massive backlash against AI is also entirely plausible (could be for environmental reason, social reason, or you can even use that old reliable staple, the Machine War.)

[Moose-tache]

Somewhere back in 2021, there's a conworlder asking "I want my future conworld to not be dominated by NFTs. How can I possibly achieve this?"

We are not living in a general-artificial-intelligence revolution. We are living in a time when Markov chains are just convincing enough that Capital can fool itself into thinking it can finally build the workerless treats machine it's always wanted. This, too, shall pass. In five years, you'll feel very silly for thinking you have to do anything special to work around "the AI future."

[malloc]

Somewhere back in 2021, there's a conworlder asking "I want my future conworld to not be dominated by NFTs. How can I possibly achieve this?"

We are not living in a general-artificial-intelligence revolution. We are living in a time when Markov chains are just convincing enough that Capital can fool itself into thinking it can finally build the workerless treats machine it's always wanted. This, too, shall pass. In five years, you'll feel very silly for thinking you have to do anything special to work around "the AI future."

What about all the claims that LLMs are proving and refuting conjectures that have stumped mathematicians for decades? That sounds like intelligence to me, even if not quite general. Certainly it's more than just a Markov chain.

[jal]

What about all the claims that LLMs are proving and refuting conjectures that have stumped mathematicians for decades? That sounds like intelligence to me, even if not quite general. Certainly it's more than just a Markov chain.

Those claims are highly overrated, and though I'm not certain whether modern LLMs employ Markov chains, it's still the result of token prediction. AIs spotting tumors on CT scans also aren't "intelligent" by any definition, though they can do things humans can't.


JAL

[WeepingElf]

What about all the claims that LLMs are proving and refuting conjectures that have stumped mathematicians for decades? That sounds like intelligence to me, even if not quite general. Certainly it's more than just a Markov chain.

Those claims are highly overrated, and though I'm not certain whether modern LLMs employ Markov chains, it's still the result of token prediction. AIs spotting tumors on CT scans also aren't "intelligent" by any definition, though they can do things humans can't.

I don't know much about LLM technology, but from what little I understand, they are more sophisticated than Markov chains - but they are still much closer to Markov chains than to human minds, I'd guess. As you say, it's all about token prediction, not about true intelligence. Doesn't Latin intelligere mean 'to understand'? LLMs don't understand anything; they are only quite good at predicting which token follows next.

[Moose-tache]

"But the abacus has allowed Mathematics to do things no unaided human has done before! Surely we will all be replaced by abacuses in the next few years!"

I will concede that LLMs have some use case somewhere as a tool. They are incremental improvements of existing technology. And calling them MArkov chains was meant to be slightly hyperbolic. But my point still stands. What they call "AI" is hyped up by people who have Capital-related reasons to hype it up, not because of any real capacity that they have.

[Richard W]

"But the abacus has allowed Mathematics to do things no unaided human has done before! Surely we will all be replaced by abacuses in the next few years!"

Computer is no longer a human occupation.

[Travis B.]

"But the abacus has allowed Mathematics to do things no unaided human has done before! Surely we will all be replaced by abacuses in the next few years!"

Computer is no longer a human occupation.

Most people today do not realize that computer once referred to a human job.

[Imralu]

"But the abacus has allowed Mathematics to do things no unaided human has done before! Surely we will all be replaced by abacuses in the next few years!"

Computer is no longer a human occupation.

Most people today do not realize that computer once referred to a human job.

Now it is time for beach to be an occupation.