Yet wealth, as measured by anything of your choosing (juliansimon.com) lifespan, general health, cost of food, cost of materials, keeps getting better and better.
This whole kite thing is getting stupid. They had metalworking, too, so there's no reason they couldn't have built a steam engine, giant crane, and hydraulics while they were at it, and just heaved the stones into place. All they needed were some clever engineers, then they all conveniently forget all that knowledge without writing any of it down.
It's been pointed out that modern, charlatanistic booksellers making claims about ancient aliens doing all the hard work does a real disservice to the ingenuity of ancient engineers.
These are the same guys who will claim NASA murdered 10 people to keep the secret of the "faked" moon landing.
Chimps and gorillas don't "bark" at themselves in the mirror, so to speak, if that's what you mean, and react differently than they would if, say, they saw another gorilla thru a glass pane. I think so, don't have any links either.:(
> Subjective perceptual experience is a private
> experience and can neither be verified nor
> falsified.
By the way, this has been more or less enshrined in philosophy as a Truth. I view it as an engineering and physics problem. Something generates this, therefore there will be a way to detect it. We just haven't figured it out, yet.
(Note also that claiming dualism [including "hard" emergent properties] only pushes the physics off to another plane of reality, so to speak. Just how does this "spirit stuff" operate? Let's see now...)
Exactly! One would think multiple giant kites lifting many hundred-ton or thousand-ton stones into place would have been at least marginally spectacular, causing at least one wall painting somewhere. It's spectacular in today's world of jets and spacecraft. Back then, pre-planes, it would have been such a wonder that people might really have viewed Egyptian kings as gods.
Just because it can be done in such and such a way using modern technologies, but with ancient techniques and materials doesn't mean it was done that way. The giant, steam-powered mechanical iron spider in Wild Wild West shows that much...
Old Geezer: TRON was a command in TRS-80 Basic, and probably other systems before then, and stood for TRace ON, where line numbers would be printed as the program executed, for debugging purposes.
Won't have to. It will be obvious to it with a rudimentary physics database that such a plan would be hideously inefficient, especially in a light-poor world.
I wouldn't. You'd quickly end up in a robot-controlled super-socialism, where everyone was kept in a cage and kept unconscious (lest they struggle) and fed intravenously.
Then you'd have to start the Star-Trekky crap like "being caged harms us."
My favorite way of looking at the Church-Turing thesis is this way:
We have this general concept of a most-powerful computational machine that can perform at most a finite number of calculations in a finite period of time. The Turing machine, they suggest, is this machine. It hasn't been proven, but no more powerful finite machine has been invented yet.
Now, for AI, a Turing machine could theoretically simulate all known physics to any arbitrary degree of accuracy, even for the entire universe, random processes included, from beginning to well in the future. (From which one immediately wonders what metaphysics generates the randomness of quantum events.)
So, any doubter that a brain is no more powerful than a Turing machine, computationally, would have to point out exactly where the simulation would be forced to part from reality. Quantum uncertainty? Easily simulated. Quantum randomness? Easily simulated, and it's a stretch to argue it has anything to do with the subjective perceptual experiance (people do, but because it's their last straw, an argument by omission of all other possibilities.) Unknown physics? It will be known someday as long as it can be detected, and if it can't, you don't now it's there, now do you?
The only real escape is the Searle argument that there is something about the universe we don't understand, and that is how the subjective perceptual experience arises. When that is figured out, that could be simulated, too, most likely. If it can't, it violates the Church-Turing thesis, and would be extremely interesting in its own right.
The point is though, that as a thought experiment, something would have to happen when you turned it on. What?
Would it sit there, unfunctioning? Would it scream bloody murder as you simulated an input of a pin through its finger? If so, would it actually perceive pain as a human does, or would it just act like it because the information processing "brain" in it is identical to a human's, but that is separate from perceptual experiences?
Something creates this experience, and if we could figure out how that physics works, we would know which of those three ways such a simulation would work (not to mention being able to better grasp the ethics.)
Exactly. The problem is that it's not all that obvious that humans don't work this way. The Chinese room demonstrates that the AI philosophy concept of "deep understanding" may be a bogus one.
For another example, consider your "deep understanding" of addition. A computer, obviously, only "understands" a set of mechanical rules to transform one set of symbols (a pattern) into another pattern. But a human clearly has a deeper understanding of what's going on.
...or do they? You know that 3 + 4 is 7 because of deep understanding (not having to calculate it) or because you've memorized it? You know that 600 / 2 is 300 because you've got deep understanding? Or because you've memorized 6 / 2 is 3 and the 0's won't change in that situation?
The reason Cyc was created in the first place was because AI researchers threw up their hands and guessed the only way to get a massive amount of data into an AI might be to just build it by hand. They do that, and turn a reasoning engine loose on it. There would be no fancy shortcuts. If symbol manipulation was what intelligence was based on, then you'd just have to keep plowing away ontil VGer woke up.
> Unfortunately, the Turing test is no indicator
> of strong AI at all, just a very good rules
> system.
Not true. The Turing test's point was not to be a practical way to test for AI, but to point out that we have no way to know if something is truly intelligent or not; the best we can do is interact with it.
When physics gets around to trying to answer what is it in reality that generates the subjective perceptual experience of consciousness, then and only then will we be able to move beyond the Turing test to see if something truly has that type of consciousness. (Self-awareness isn't it. Animals don't have any real self-awareness (pre-reflective cogitos, they) yet they almost certainly have subjective perceptual experiences. And why? Because their "machinery" is similar to ours and their behavior is similar, and from that we draw the conclusion, just like each of us does that other humans are that way too.)
> Searle proved years ago that no rules-based
> system (ie. a Turing machine) can ever be
> truly intelligent, no matter how much so it
> seems.
No, he suggested (not proved) that the subjective perceptual experience was not information-derived in nature, but rather a natural feature of the universe and reality. (Even pushing it to dualism, i.e. a spiritual world/plan, only shoves the physics off to another type of reality.)
He pointed out that you could, in principal, build a machine (buckets and water pipes one of his best examples) that would exactly mimic the functioning of a human brain. The great question, of course, is what would happen next? Would it respond exactly like a human, but be devoid of the subjective perceptual experience (interesting, it means thinking and memories are separate from "you".) Would it not respond at all, meaning thinking requires whatever universal machinery that also happens to spawn the subjective perceptual experience? Or would it respond like a human, and actually have subjective perceptual experiences? Searle claimed no to the last question, many other AI people claim yes.
Until physics gets around to it, though, no one will really know. Would a simulation of a brain have that property? There are machines that do; the brain is one of them. Other machines might, too, but he claimed a simple duplication of information activity in the brain would not give rise to the subjective experience; only the peculiar and still unknown physical properties of the brain, its cells, its atoms, whatever, do.
It's not adding content. I sure don't want the government telling me I can't have a machine that will give me definitions for words on a page, encyclopedic descriptions of things, and so on. That Microsoft (or any other company or organization or government) might want to supply me with such a machine, and that said machine might slant things their way the way browser blockers hide criticism of browser blockers in addition to pr0n is a risk I'd gladly take over having it be prevented by threat.
Slashdot Mirror
Yet wealth, as measured by anything of your choosing (juliansimon.com) lifespan, general health, cost of food, cost of materials, keeps getting better and better.
Golly!
This whole kite thing is getting stupid. They had metalworking, too, so there's no reason they couldn't have built a steam engine, giant crane, and hydraulics while they were at it, and just heaved the stones into place. All they needed were some clever engineers, then they all conveniently forget all that knowledge without writing any of it down.
It's been pointed out that modern, charlatanistic booksellers making claims about ancient aliens doing all the hard work does a real disservice to the ingenuity of ancient engineers.
These are the same guys who will claim NASA murdered 10 people to keep the secret of the "faked" moon landing.
Chimps and gorillas don't "bark" at themselves in the mirror, so to speak, if that's what you mean, and react differently than they would if, say, they saw another gorilla thru a glass pane. I think so, don't have any links either. :(
> Subjective perceptual experience is a private
> experience and can neither be verified nor
> falsified.
By the way, this has been more or less enshrined in philosophy as a Truth. I view it as an engineering and physics problem. Something generates this, therefore there will be a way to detect it. We just haven't figured it out, yet.
(Note also that claiming dualism [including "hard" emergent properties] only pushes the physics off to another plane of reality, so to speak. Just how does this "spirit stuff" operate? Let's see now...)
Exactly! One would think multiple giant kites lifting many hundred-ton or thousand-ton stones into place would have been at least marginally spectacular, causing at least one wall painting somewhere. It's spectacular in today's world of jets and spacecraft. Back then, pre-planes, it would have been such a wonder that people might really have viewed Egyptian kings as gods. Just because it can be done in such and such a way using modern technologies, but with ancient techniques and materials doesn't mean it was done that way. The giant, steam-powered mechanical iron spider in Wild Wild West shows that much...
Old Geezer: TRON was a command in TRS-80 Basic, and probably other systems before then, and stood for TRace ON, where line numbers would be printed as the program executed, for debugging purposes.
Won't have to. It will be obvious to it with a rudimentary physics database that such a plan would be hideously inefficient, especially in a light-poor world.
I wouldn't. You'd quickly end up in a robot-controlled super-socialism, where everyone was kept in a cage and kept unconscious (lest they struggle) and fed intravenously.
Then you'd have to start the Star-Trekky crap like "being caged harms us."
Why isn't it deterministic? Did they include a seedless random number generator?
Someone really should post that link to the site where someone turned Eliza loose inside a sex chatroom...
If I recall, at one point it asked something like "If I turn around [can't see somthing] is it still there?"
Whatever else, it pointed out some extremely interesting "observations" of things we take for granted in our world views.
Not to mention "ashes to ashes, dust to dust", which is probably (!) in there somewhere. All much more poetic than a "lump of sand". Sheesh.
I think it was Curly who first said that.
It's a meme that can ONLY spread by cut-and-paste. I wonder how that slows its spread, if not killing it outright.
But your brain is just a huge list of rules pushing symbols around. They are your neurons with chemically weighted connections pushing impulses.
Hmmmm, something else must be going on here. I wonder what it is?
My favorite way of looking at the Church-Turing thesis is this way:
We have this general concept of a most-powerful computational machine that can perform at most a finite number of calculations in a finite period of time. The Turing machine, they suggest, is this machine. It hasn't been proven, but no more powerful finite machine has been invented yet.
Now, for AI, a Turing machine could theoretically simulate all known physics to any arbitrary degree of accuracy, even for the entire universe, random processes included, from beginning to well in the future. (From which one immediately wonders what metaphysics generates the randomness of quantum events.)
So, any doubter that a brain is no more powerful than a Turing machine, computationally, would have to point out exactly where the simulation would be forced to part from reality. Quantum uncertainty? Easily simulated. Quantum randomness? Easily simulated, and it's a stretch to argue it has anything to do with the subjective perceptual experiance (people do, but because it's their last straw, an argument by omission of all other possibilities.) Unknown physics? It will be known someday as long as it can be detected, and if it can't, you don't now it's there, now do you?
The only real escape is the Searle argument that there is something about the universe we don't understand, and that is how the subjective perceptual experience arises. When that is figured out, that could be simulated, too, most likely. If it can't, it violates the Church-Turing thesis, and would be extremely interesting in its own right.
The point is though, that as a thought experiment, something would have to happen when you turned it on. What?
Would it sit there, unfunctioning? Would it scream bloody murder as you simulated an input of a pin through its finger? If so, would it actually perceive pain as a human does, or would it just act like it because the information processing "brain" in it is identical to a human's, but that is separate from perceptual experiences?
Something creates this experience, and if we could figure out how that physics works, we would know which of those three ways such a simulation would work (not to mention being able to better grasp the ethics.)
Exactly. The problem is that it's not all that obvious that humans don't work this way. The Chinese room demonstrates that the AI philosophy concept of "deep understanding" may be a bogus one.
For another example, consider your "deep understanding" of addition. A computer, obviously, only "understands" a set of mechanical rules to transform one set of symbols (a pattern) into another pattern. But a human clearly has a deeper understanding of what's going on.
...or do they? You know that 3 + 4 is 7 because of deep understanding (not having to calculate it) or because you've memorized it? You know that 600 / 2 is 300 because you've got deep understanding? Or because you've memorized 6 / 2 is 3 and the 0's won't change in that situation?
People are dismissing Cyc far too out-of-hand.
The reason Cyc was created in the first place was because AI researchers threw up their hands and guessed the only way to get a massive amount of data into an AI might be to just build it by hand. They do that, and turn a reasoning engine loose on it. There would be no fancy shortcuts. If symbol manipulation was what intelligence was based on, then you'd just have to keep plowing away ontil VGer woke up.
> Unfortunately, the Turing test is no indicator
> of strong AI at all, just a very good rules
> system.
Not true. The Turing test's point was not to be a practical way to test for AI, but to point out that we have no way to know if something is truly intelligent or not; the best we can do is interact with it.
When physics gets around to trying to answer what is it in reality that generates the subjective perceptual experience of consciousness, then and only then will we be able to move beyond the Turing test to see if something truly has that type of consciousness. (Self-awareness isn't it. Animals don't have any real self-awareness (pre-reflective cogitos, they) yet they almost certainly have subjective perceptual experiences. And why? Because their "machinery" is similar to ours and their behavior is similar, and from that we draw the conclusion, just like each of us does that other humans are that way too.)
> Searle proved years ago that no rules-based
> system (ie. a Turing machine) can ever be
> truly intelligent, no matter how much so it
> seems.
No, he suggested (not proved) that the subjective perceptual experience was not information-derived in nature, but rather a natural feature of the universe and reality. (Even pushing it to dualism, i.e. a spiritual world/plan, only shoves the physics off to another type of reality.)
He pointed out that you could, in principal, build a machine (buckets and water pipes one of his best examples) that would exactly mimic the functioning of a human brain. The great question, of course, is what would happen next? Would it respond exactly like a human, but be devoid of the subjective perceptual experience (interesting, it means thinking and memories are separate from "you".) Would it not respond at all, meaning thinking requires whatever universal machinery that also happens to spawn the subjective perceptual experience? Or would it respond like a human, and actually have subjective perceptual experiences? Searle claimed no to the last question, many other AI people claim yes.
Until physics gets around to it, though, no one will really know. Would a simulation of a brain have that property? There are machines that do; the brain is one of them. Other machines might, too, but he claimed a simple duplication of information activity in the brain would not give rise to the subjective experience; only the peculiar and still unknown physical properties of the brain, its cells, its atoms, whatever, do.
Ehh, let 'em land in the Detroit area, with 2.5 guns per capita, and see how well a few thousand troops in a few hundred helicopters do.
> dictatorships shouldn't be able to vote on many (most?) items.
Well that would rule out democracies, since they're dictatorships of the majority.
It's not adding content. I sure don't want the government telling me I can't have a machine that will give me definitions for words on a page, encyclopedic descriptions of things, and so on. That Microsoft (or any other company or organization or government) might want to supply me with such a machine, and that said machine might slant things their way the way browser blockers hide criticism of browser blockers in addition to pr0n is a risk I'd gladly take over having it be prevented by threat.
And big, old car engines were built for you, too. You can take yours out, rebore it, rebuild the engine, change the timing.
Meanwhile, 99.99% of the population goes the other way.