Zompist Bboard Again

zompist wrote: ↑Sat Jun 07, 2025 8:06 pmAnd you're an expert in computation, human neurology, and philosophy of mind from what experience exactly?

I've been following AI for longer than you've been alive. AI researchers can be optimists, but few of them are as credulous about AI, or as dismissive of human brains, as you are.

There are entire corporations who have staked their futures and billions in investments on AI achieving or exceeding human-level intelligence. There are plenty of CEOs and investors who fully expect some radical advances in AI including general intelligence in mere years. Would you consider vast swathes of the tech industry only a credulous few?

Not the same claim at all. People have poured money into cold fusion and perpetual motion, too. You don't seem to understand when a problem is hard.

Except that the laws of physics prohibit cold fusion and perpetual motion whereas human-level intelligence clearly exists. The people trying to create cold fusion were dismissing the entire physics community whereas AI researchers are merely trying to replicate something that already has billions of functioning examples. We already know that human intelligence arises from the physical brain rather than some Cartesian soul and we have have computers capable of simulating vast networks of neurons.

malloc wrote: ↑Sat Jun 07, 2025 9:00 pm

zompist wrote: ↑Sat Jun 07, 2025 8:06 pmAnd you're an expert in computation, human neurology, and philosophy of mind from what experience exactly?

please answer the question, Malloc.

I've been following AI for longer than you've been alive. AI researchers can be optimists, but few of them are as credulous about AI, or as dismissive of human brains, as you are.

There are entire corporations who have staked their futures and billions in investments on AI achieving or exceeding human-level intelligence.

There were entire corporations who did exactly that with cold fusion and perpetual motion.

Not the same claim at all. People have poured money into cold fusion and perpetual motion, too. You don't seem to understand when a problem is hard.

Except that the laws of physics prohibit cold fusion and perpetual motion whereas human-level intelligence clearly exists.

Humans clearly exist. Machines with human-level intelligence do not clearly exist, and have yet to be proven to be anything other than another attempt to make castles in the sky, like perpetual motion.

and we have have computers capable of simulating vast networks of neurons.

This is one of the reasons I think you are an AI, Malloc. Its because you don't accept new input, no matter how many times it is placed before you.

keenir wrote: ↑Sat Jun 07, 2025 9:48 pm

malloc wrote: ↑Sat Jun 07, 2025 9:00 pm

zompist wrote: ↑Sat Jun 07, 2025 8:06 pmAnd you're an expert in computation, human neurology, and philosophy of mind from what experience exactly?

please answer the question, Malloc.

No, I am not an expert but I have read about all of those things. The experts in neurology and philosophy of mind overwhelmingly agree that human intelligence arises from the physical brain rather than some metaphysical substance that machines cannot replicate. Nobody takes Cartesian dualism seriously anymore, certainly not in neuroscience.

There were entire corporations who did exactly that with cold fusion and perpetual motion.

There were never multiple trillion-dollar corporations dedicated to developing and marketing cold fusion. The concept never advanced beyond a handful of scientists in the late 80s with the vast majority of nuclear physicists soundly rejecting cold fusion.

Humans clearly exist. Machines with human-level intelligence do not clearly exist, and have yet to be proven to be anything other than another attempt to make castles in the sky, like perpetual motion.

Your objection implies that the laws of physics prohibit machines from achieving general intelligence just as they prevent perpetual motion. That would require something about human cognition that transcends physics and thus the capabilities of engineering, something which nobody in neurology or philosophy of mind accepts.

malloc wrote: ↑Sat Jun 07, 2025 9:00 pm

zompist wrote: ↑Sat Jun 07, 2025 8:06 pmAnd you're an expert in computation, human neurology, and philosophy of mind from what experience exactly?

I've been following AI for longer than you've been alive. AI researchers can be optimists, but few of them are as credulous about AI, or as dismissive of human brains, as you are.

There are entire corporations who have staked their futures and billions in investments on AI achieving or exceeding human-level intelligence. There are plenty of CEOs and investors who fully expect some radical advances in AI including general intelligence in mere years. Would you consider vast swathes of the tech industry only a credulous few?

Do you understand how capitalism works? "Vast swatches of the tech industry" come down to about ten CEOs-- most of which made their money elsewhere and can burn it on AI. Remember the "Metaverse" that Mark Zuckerberg was on about? $45 billion spent, and the project is basically a failure. Why do you worship these people?

Do you understand how overvalued AI is? OpenAI has a market valuation of $300 billion, on revenues of $4 billion. The ratio is 75 to 1; compare Microsoft (13 to 1) or Walmart (1.15 to 1).

Not the same claim at all. People have poured money into cold fusion and perpetual motion, too. You don't seem to understand when a problem is hard.

Except that the laws of physics prohibit cold fusion and perpetual motion whereas human-level intelligence clearly exists. The people trying to create cold fusion were dismissing the entire physics community whereas AI researchers are merely trying to replicate something that already has billions of functioning examples.

You don't understand when a problem is hard. I gave you another example: settling Mars. It's physically possible; it's also not going to happen soon.

Or take understanding fundamental physics. We're basically using the same model that was finalized in the 1970s, fifty years ago. New theories like superstring theory haven't worked out, much less replaced it. We know a lot more about things the theory doesn't handle (i.e. dark matter, dark energy), but we don't have a theory underlying them. We don't have a unification of quantum mechanics and gravity after a hundred years of trying.

But you read some huckster's prediction of "AGI" next year and you swallow it. Good thing you have no money to lose on it.

keenir wrote: ↑Sat Jun 07, 2025 9:48 pm

malloc wrote: ↑Sat Jun 07, 2025 9:00 pm

please answer the question, Malloc.

No, I am not an expert but I have read about all of those things. [/quote]

and I've read how to make bologna; you want me to send you some of the food I've prepared? if you say Yes, you have to eat it.

There were entire corporations who did exactly that with cold fusion and perpetual motion.

There were never multiple trillion-dollar corporations dedicated to developing and marketing cold fusion.

none that you noticed, clearly; yet you salivate at the mere mention of any of the rich techbros' commenting about AI.

Humans clearly exist. Machines with human-level intelligence do not clearly exist, and have yet to be proven to be anything other than another attempt to make castles in the sky, like perpetual motion.

Your objection implies that the laws of physics prohibit machines from achieving general intelligence just as they prevent perpetual motion.

no, I was rebutting your claim that, because humans exist, anything and everything can develop human-level intelligence.

we're currently around 14 or 15 billion years into this current universe...care to point to where your sapient machines are proving you right? its okay if they are Berserkers.

That would require something about human cognition that transcends physics and thus the capabilities of engineering, something which nobody in neurology or philosophy of mind accepts.

it would only require something, if you are correct.

Did any of you read the story "The Beautiful and the Sublime" by Bruce Sterling?

Some cold water poured on the claim that recent LLMS can "do reasoning":

https://pivot-to-ai.com/2025/06/08/appl ... -to-think/

Oh but they pass benchmarks! Nah, those are rigged and not reported honestly.

zompist wrote: ↑Sat Jun 07, 2025 10:36 pmDo you understand how capitalism works? "Vast swatches of the tech industry" come down to about ten CEOs-- most of which made their money elsewhere and can burn it on AI. Remember the "Metaverse" that Mark Zuckerberg was on about? $45 billion spent, and the project is basically a failure. Why do you worship these people?

Do you understand how overvalued AI is? OpenAI has a market valuation of $300 billion, on revenues of $4 billion. The ratio is 75 to 1; compare Microsoft (13 to 1) or Walmart (1.15 to 1).

Sure but these corporations (and the field of AI research more generally) include thousands of scientists and engineers who actually develop the AI models for a living. Are all those people just cynics who secretly consider the C-suite delusional and find the whole project of artificial general intelligence preposterous?

But you read some huckster's prediction of "AGI" next year and you swallow it. Good thing you have no money to lose on it.

I have never claimed that it will happen next year, but I could certainly see it happening within my lifetime. All the pieces are there, from simulated neurons to billions in funding every year. All that remains is figuring out how to connect the neurons in a way that approximates the structure of biological brains. That may sound quite daunting but you never know when the next breakthrough in neuroscience will give us the key.

malloc wrote: ↑Sun Jun 08, 2025 7:47 pm

zompist wrote: ↑Sat Jun 07, 2025 10:36 pmDo you understand how capitalism works? "Vast swatches of the tech industry" come down to about ten CEOs-- most of which made their money elsewhere and can burn it on AI. Remember the "Metaverse" that Mark Zuckerberg was on about? $45 billion spent, and the project is basically a failure. Why do you worship these people?

Do you understand how overvalued AI is? OpenAI has a market valuation of $300 billion, on revenues of $4 billion. The ratio is 75 to 1; compare Microsoft (13 to 1) or Walmart (1.15 to 1).

Sure but these corporations (and the field of AI research more generally) include thousands of scientists and engineers who actually develop the AI models for a living. Are all those people just cynics who secretly consider the C-suite delusional and find the whole project of artificial general intelligence preposterous?

You've evidently never worked as a programmer. Do programmers find their CEOs delusional? Yes, most of the time. But when the median wage is $130K you can put up with a lot.

But the real problem here is your weaselling about what the "thousands of scientists and engineers" are doing. Are they making "AGI"? No, that's the marketing sludge that you can't see through because you want to scare yourself. What they're doing is making LLMs, or other forms of generative "AI", things which no one but yourself mistakes for AGI. And honestly, besides being well paid, it's probably interesting work.

I have never claimed that it will happen next year, but I could certainly see it happening within my lifetime. All the pieces are there, from simulated neurons to billions in funding every year. All that remains is figuring out how to connect the neurons in a way that approximates the structure of biological brains. That may sound quite daunting but you never know when the next breakthrough in neuroscience will give us the key.

None of this addresses the fact that you have no idea what a hard problem is.

zompist wrote: ↑Sun Jun 08, 2025 8:00 pmBut the real problem here is your weaselling about what the "thousands of scientists and engineers" are doing. Are they making "AGI"? No, that's the marketing sludge that you can't see through because you want to scare yourself. What they're doing is making LLMs, or other forms of generative "AI", things which no one but yourself mistakes for AGI. And honestly, besides being well paid, it's probably interesting work.

So nobody is actually researching AGI itself, just LLMs currently?

For that matter, what really distinguishes neural nets in computing from biological brains that would make emulating the latter with the former such a hard problem?

malloc wrote: ↑Sun Jun 08, 2025 8:08 pm For that matter, what really distinguishes neural nets in computing from biological brains that would make emulating the latter with the former such a hard problem?

okay, try this: what is the difference between a handful of nerve cells grown in a petri dish, from a human brain in a skull? if you have lots of petri dishes full of nerve cells, do you suddenly have a human level intelligence?

malloc wrote: ↑Sun Jun 08, 2025 7:47 pm

zompist wrote: ↑Sat Jun 07, 2025 10:36 pmDo you understand how capitalism works? "Vast swatches of the tech industry" come down to about ten CEOs-- most of which made their money elsewhere and can burn it on AI. Remember the "Metaverse" that Mark Zuckerberg was on about? $45 billion spent, and the project is basically a failure. Why do you worship these people?

Do you understand how overvalued AI is? OpenAI has a market valuation of $300 billion, on revenues of $4 billion. The ratio is 75 to 1; compare Microsoft (13 to 1) or Walmart (1.15 to 1).

Sure but these corporations (and the field of AI research more generally) include thousands of scientists and engineers who actually develop the AI models for a living.

the fact that there are these hundreds of corporations, each of which has thousands of scientists and engineers hard at work on developing AI/AI Models...and that we still are light-years from a real AI, that should tell you something.

But you read some huckster's prediction of "AGI" next year and you swallow it. Good thing you have no money to lose on it.

I have never claimed that it will happen next year,

you keep saying we're on the doorstep of it happening, that it'll happen in a month or so, etc.

but I could certainly see it happening within my lifetime. All the pieces are there, from simulated neurons to billions in funding every year.

the phrase "throwing good money after bad" comes to mnid.

All that remains is figuring out how to connect the neurons in a way that approximates the structure of biological brains.

see above with the dishes.

That may sound quite daunting but you never know when the next breakthrough in neuroscience will give us the key.

if you don't know what the needed breakthrough is, though, it becomes harder to aim.

malloc wrote: ↑Sat Jun 07, 2025 6:29 pm

Torco wrote: ↑Sat May 31, 2025 3:50 pmThey indeed don't know a lot about it. then again, no one is trying to make a human-like mind -that i know of- so it's even less likely that they will do it.

Except that multiple massive corporations have specifically put forth artificial general intelligence as one of their goals. You can dispute the feasibility of this goal but you cannot claim that nobody is trying to replicate the abilities of the human mind.

but AGI is still not a person. just some nebulous concept of having a model that's capable of adequately performing functions that are very broad in scope, about as broad as a person, or something. you fall again and again into this congnitive trap of "people can do X,Y,Z. software is doing X, and they're working on Y and Z >>> therefore, they're making software people". but no, it's not software people, not anymore than the petrol engine is a metal horse. the engine is enough of a metal horse that it dramatically decreased the horse population, but it's not enough of a metal horse that you can to any significant degree understand it by analogy with the horse.

put even more simply, to do what a horse can do is not the same as to be a horse, or a superior-horse. to do what a mind can do is not to be a mind, or a superior-mind. function is not structure. ability is not essence.

malloc wrote: ↑Sat Jun 07, 2025 6:29 pm Except that multiple massive corporations have specifically put forth artificial general intelligence as one of their goals. You can dispute the feasibility of this goal but you cannot claim that nobody is trying to replicate the abilities of the human mind.

see here you're closer to truth: people are trying to replicate concrete abilities of the human mind, and for some of them succeeding! but functional replication is not replication itself: i have a hot plate next to me [nice little bit of tech] which replicates some functions of a wood fire, i.e. heating up my morning milk. it is not, however, a wood fire.

keenir wrote: ↑Sat Jun 07, 2025 8:27 pmMalloc, people have been trying to replicate the human mind since the days of Byron, if not Aristotle.

in fairness, this is going too far [lmao am i the centrist here?]. they've been trying since aristotle, sure, but they haven't been getting much success at the functional part it till relatively recently. AGI is *possible*, if not stricto sensu, in a sensu stricto enough to have the powers that be decide we're all not necessary. it's not silly to worry about the negative effects of these algorithms: what's silly is to think about the in the particular way malloc does.

Torco wrote: ↑Mon Jun 09, 2025 9:05 am

keenir wrote: ↑Sat Jun 07, 2025 8:27 pmMalloc, people have been trying to replicate the human mind since the days of Byron, if not Aristotle.

in fairness, this is going too far [lmao am i the centrist here?]. they've been trying since aristotle, sure, but they haven't been getting much success at the functional part it till relatively recently. AGI is *possible*, if not stricto sensu, in a sensu stricto enough to have the powers that be decide we're all not necessary. it's not silly to worry about the negative effects of these algorithms: what's silly is to think about the in the particular way malloc does.

If I remember correctly, my saying "since the days of Byron, if not Aristotle" was a response to Malloc's statement of - in essence - people are trying to create human-like minds, so it'll be attained any decade now...so I was trying to disprove the "any day now" by pointing out how long the "people are trying" has been going on: far more than the few decades Malloc may've thought it had.

I'm probably wrong, but I suspect that the "our creation will kill us all" is a recent innovation -- tales of Babbage Engines, Galatea, and Frankenstein, none of those were ever thought capable of the sort of thing that Skynet tried.

mmmm i wouldn't be so sure. premodern peoples, some of 'em anyway, seem to be -see graber and wrenwrow- acutely aware of how the implementation of this or that social system affects life. and i think it's AI-as-a-social-fact that people are ultimately scared of, not AI-as-sheer-tech

keenir wrote: ↑Mon Jun 09, 2025 9:13 am tales of Babbage Engines,

Serious question: How much was the Babbage Engine talked about in its day? I had the impression that part of the reason why it was eventually abandoned was that few people saw much potential in it at the time, and that then, it was mostly forgotten for a while, until historians of technology in the late 20th Century discovered to their slight bemusement that someone had built a forerunner of the modern computer in the Victorian Age.

Is my impression wrong?

Who's going to tell malloc that there are various animals that are better at specific tasks than humans? Maybe we can have him fearing the inevitable zebra or chimp uprisings if we play it right. /s

linguistcat wrote: ↑Mon Jun 09, 2025 10:40 am Who's going to tell malloc that there are various animals that are better at specific tasks than humans? Maybe we can have him fearing the inevitable zebra or chimp uprisings if we play it right. /s

I dunno, perhaps I'm misremembering something, but didn't that issue already come up at some point in the history of this thread? (Not interested in re-reading 77 pages now.)

Raphael wrote: ↑Mon Jun 09, 2025 10:50 am

linguistcat wrote: ↑Mon Jun 09, 2025 10:40 am Who's going to tell malloc that there are various animals that are better at specific tasks than humans? Maybe we can have him fearing the inevitable zebra or chimp uprisings if we play it right. /s

I dunno, perhaps I'm misremembering something, but didn't that issue already come up at some point in the history of this thread? (Not interested in re-reading 77 pages now.)

Ah. Unfortunately I've only been skimming along so I must have missed it if it had come up.

Raphael wrote: ↑Mon Jun 09, 2025 10:50 am

linguistcat wrote: ↑Mon Jun 09, 2025 10:40 am Who's going to tell malloc that there are various animals that are better at specific tasks than humans? Maybe we can have him fearing the inevitable zebra or chimp uprisings if we play it right. /s

you mean Planet of the Apes isn't a documentary?

I dunno, perhaps I'm misremembering something, but didn't that issue already come up at some point in the history of this thread? (Not interested in re-reading 77 pages now.)

I don't think it did...if we approached that topic, it was animal senses (hawks can see better, etc)

keenir wrote: ↑Sun Jun 08, 2025 10:01 pmokay, try this: what is the difference between a handful of nerve cells grown in a petri dish, from a human brain in a skull? if you have lots of petri dishes full of nerve cells, do you suddenly have a human level intelligence?

Sure but neural nets are not simply random clusters of artificial neurons strewn around in small containers. There are well-developed techniques and principles for connecting the neurons in useful ways. Currently we have not yet figured out how to connect them just like the human brain, but we have figured out how to create neural nets that emulate numerous cognitive tasks once reserved for humans. Even without replicating the connectome of the human brain, neural nets are remarkably effective.

the fact that there are these hundreds of corporations, each of which has thousands of scientists and engineers hard at work on developing AI/AI Models...and that we still are light-years from a real AI, that should tell you something.

What makes you so sure that we are light-years away from true AI, though? Several years ago everyone assumed that generative AI capable of drawing pictures and writing novels was quite a long way off. Obviously that doesn't mean that AGI will happen tomorrow, but it seems premature to declare it light-years away. Considering the astonishing progress in AI over the past few years, I feel we have already traversed considerable parsecs after all.

Also who the hell are all these guests on the site and are they the reason it keeps crashing?