It follows that some amount of intelligent application of the theories that we know to be true may be able to replicate it many orders of magnitude faster.
Yet nothing of the sort has happened so far. Oh, and incidentally, the "theories that we know to be true" do not include Quantum Mechanics at this time. It is just very well verified, but it does not go together with Gravity, which is also very well verified. So one or both must be wrong and nobody so far has a clue which it is. And one thing we do know is that the brain is a heavy user of quantum effects.
He does not need that. He just points out it cannot be ruled out. The unproven assumption "there is nothing magic" is from you. Got any proof? Because the last results from Neuroscience say "the closer we look, the more mysterious things become".
Incidentally, physics as known today is known to be incomplete and partially wrong. If the brain uses some of the unknown parts, that is already enough to completely invalidate your assumption.
Not knowledge, however this thing can process vast amounts of knowledge and provide a response which has a very high probability of being correct and/or at the very least relevant, and it can do this in real-time.
It actually processes data, not knowledge (yes, the distinction is critical here) and it cannot actually do this in real-time. It takes significant pre-computation time to get there and only the last step is somewhat real-time. But that is not the main problem. It unfortunately has no "very high probability of being correct" (as it does not do any fact-checking and there are countless things the press consistently get wrong) and being "relevant" without being correct is worse than having nothing. This is were anti-vaxxers and flat-earthers come from. Their arguments are most decidedly relevant, but they are deeply wrong.
Well, yes. But in the explanation part you get things like Quantum Mechanics which are truly bizarre and very different from what you would expect. By some definition some effects in Quantum Mechanics do qualify as "magic".
But you are misunderstanding. I am not claiming that "magic" explains general intelligence. I am claiming, and with a sound scientific stance, that we do not know at all how general intelligence works and hence "magic" cannot be ruled out. It looks pretty certain that known physics will be not sufficient to describe it and it is hence unknown what extensions will be required.
Incidentally, physics is known to be incomplete and partially wrong at this time (e.g. Quantum Mechanics against Gravity) and hence an argument that a complex device where the function is not really understood and that we cannot build or analyze in detail must use known physics is pretty stupid.
Well, as there is absolutely nothing intelligent from "AI" research at this time (no, not even in the theory-stage if you require some plausibility), this is all they can do. They promised great things and at least their scientists did know they could not deliver. So they fake it and a lot of people just fall for it.
The thing is that while most humans supposedly have general intelligence, few choose to use it. (A figure of 10...15% independent thinkers pops up in actual experience, for example in advanced teaching.) To make matters worse, general intelligence is difficult to use and requires a lot of use and experience to perform well. Hence most people just go with their own mis- and preconceptions, usually copied from their peer-groups. With that, they have opinions, but no insight whatsoever.
Machines do not have "knowledge" or understanding. They have data of some level of quality. However, there is one application where the effect you describe is actually used and highly beneficial: Computer aided verification of mathematical proofs. The way this works is that a human tries to "explain" the proof to a machine. The machine then either requests more detail or verifies a step is correct. What this essentially does is aid the human in breaking down the proof into steps elementary enough to be trivial and hence accessible to automated verification.
In theory, this process could also be used to generate mathematical theorems together with their proofs automatically. However, with the computing technology and physics known, a smart, experienced and talented mathematician is more capable than a machine comprising the whole physical universe would be, so that is a non-starter.
I did that in back in school and I completely blew the other side out of the water with a claim that was obviously false. The whole thing was arranges by our pretty good Ethics teacher. I did learn that a golden tongue does not make for truth coming from it. Sadly, most of those present were just left confused and did not understand that lesson at all. Today I understand that the idea that a convincingly-sounding argument could be completely false was just too alien to them to be something they could understand.
And yes, this makes it very easy to a computer with a large database and some statistical classifier that indicated how convincing an average human would find an argument. No actual understanding of the topic required at all.
Indeed. Science can have some bias in the strength of proof required ("extraordinary facts require extraordinary proof"), but that is it. This is basically a denial-of-service protection mechanism, were an attack where some group just throws ridiculous claim after ridiculous claim and thereby prevents verification is mitigated. In the absence of facts, Science must give a strong "we do not know". In the presence of facts, it must use all of them, check them for consistency and check whether more are needed. It then gives either a conclusion or stays at "we do not know". And if the facts say something not PC or unpleasant, this is no valid reason to discount the conclusion.
That way to handle reality is the only known way to arrive at a good approximation of the truth. Yet this approach and way of thinking is completely alien to most people. They seek to have their preconceptions validated and ignore facts that would challenge these. Explains nicely why completely ridiculous ideas get so much people that "believe" them.
But debating requires better understanding of natural language than ordering pizza. To successfully rebut an argument, you need to understand its logic. I don't doubt that there's room for improvement, but it's a non-trivial step.
You misinterpret what you saw. This demo just shows that "debating" one of a topics that is previously known does not require intelligence. It also casts some real doubt on this type of competition and indicated that the "debates" done there are not actually debates in the sense most people understand that word. This machine can extract logic, but it cannot understand it. It is basically ordering of pizza on a larger scale, but with as many understanding of the nature of the action, i.e. none.
Indeed. It basically is for building cheaper (not smarter) expert systems with an natural language interface. As soon as you go off-script, this technology gets completely lost.
Slashdot Mirror
Well, that is just your personal definition. It is not shared by actual experts.
You do not understand Science one bit, really.
Since you do not even know the medical state-of-the-art here, I am going to stop responding now.
Well, I for one welcome our p-zombie friends!
It follows that some amount of intelligent application of the theories that we know to be true may be able to replicate it many orders of magnitude faster.
Yet nothing of the sort has happened so far. Oh, and incidentally, the "theories that we know to be true" do not include Quantum Mechanics at this time. It is just very well verified, but it does not go together with Gravity, which is also very well verified. So one or both must be wrong and nobody so far has a clue which it is. And one thing we do know is that the brain is a heavy user of quantum effects.
He does not need that. He just points out it cannot be ruled out. The unproven assumption "there is nothing magic" is from you. Got any proof? Because the last results from Neuroscience say "the closer we look, the more mysterious things become".
Incidentally, physics as known today is known to be incomplete and partially wrong. If the brain uses some of the unknown parts, that is already enough to completely invalidate your assumption.
Not knowledge, however this thing can process vast amounts of knowledge and provide a response which has a very high probability of being correct and/or at the very least relevant, and it can do this in real-time.
It actually processes data, not knowledge (yes, the distinction is critical here) and it cannot actually do this in real-time. It takes significant pre-computation time to get there and only the last step is somewhat real-time. But that is not the main problem. It unfortunately has no "very high probability of being correct" (as it does not do any fact-checking and there are countless things the press consistently get wrong) and being "relevant" without being correct is worse than having nothing. This is were anti-vaxxers and flat-earthers come from. Their arguments are most decidedly relevant, but they are deeply wrong.
Your requirements are pretty low. Like that of a marketeer that want to sell something at all cost.
Well, yes. But in the explanation part you get things like Quantum Mechanics which are truly bizarre and very different from what you would expect. By some definition some effects in Quantum Mechanics do qualify as "magic".
But you are misunderstanding. I am not claiming that "magic" explains general intelligence. I am claiming, and with a sound scientific stance, that we do not know at all how general intelligence works and hence "magic" cannot be ruled out. It looks pretty certain that known physics will be not sufficient to describe it and it is hence unknown what extensions will be required.
Incidentally, physics is known to be incomplete and partially wrong at this time (e.g. Quantum Mechanics against Gravity) and hence an argument that a complex device where the function is not really understood and that we cannot build or analyze in detail must use known physics is pretty stupid.
Indeed.
No, you do not get it. You are making an assumption and present it as fact. That is not Science, that is religion.
Indeed. I also like the term "imitative intelligence" (vulgo: fake intelligence).
Probably exceedingly boring. At least some creativity is required for this to work and the machine has nothing.
As IBM certainly wants those valuable defense contracts, they will make sure this does not happen...
We are finding not that machines are intelligent, but that many humans are not or at least chose not to use what they have.
Well, as there is absolutely nothing intelligent from "AI" research at this time (no, not even in the theory-stage if you require some plausibility), this is all they can do. They promised great things and at least their scientists did know they could not deliver. So they fake it and a lot of people just fall for it.
The thing is that while most humans supposedly have general intelligence, few choose to use it. (A figure of 10...15% independent thinkers pops up in actual experience, for example in advanced teaching.) To make matters worse, general intelligence is difficult to use and requires a lot of use and experience to perform well. Hence most people just go with their own mis- and preconceptions, usually copied from their peer-groups. With that, they have opinions, but no insight whatsoever.
As most "debates" these days are not about finding truth, but about "winning", the whole thing has gotten utterly corrupted.
Machines do not have "knowledge" or understanding. They have data of some level of quality. However, there is one application where the effect you describe is actually used and highly beneficial: Computer aided verification of mathematical proofs. The way this works is that a human tries to "explain" the proof to a machine. The machine then either requests more detail or verifies a step is correct. What this essentially does is aid the human in breaking down the proof into steps elementary enough to be trivial and hence accessible to automated verification.
In theory, this process could also be used to generate mathematical theorems together with their proofs automatically. However, with the computing technology and physics known, a smart, experienced and talented mathematician is more capable than a machine comprising the whole physical universe would be, so that is a non-starter.
I did that in back in school and I completely blew the other side out of the water with a claim that was obviously false. The whole thing was arranges by our pretty good Ethics teacher. I did learn that a golden tongue does not make for truth coming from it. Sadly, most of those present were just left confused and did not understand that lesson at all. Today I understand that the idea that a convincingly-sounding argument could be completely false was just too alien to them to be something they could understand.
And yes, this makes it very easy to a computer with a large database and some statistical classifier that indicated how convincing an average human would find an argument. No actual understanding of the topic required at all.
Indeed. Science can have some bias in the strength of proof required ("extraordinary facts require extraordinary proof"), but that is it. This is basically a denial-of-service protection mechanism, were an attack where some group just throws ridiculous claim after ridiculous claim and thereby prevents verification is mitigated. In the absence of facts, Science must give a strong "we do not know". In the presence of facts, it must use all of them, check them for consistency and check whether more are needed. It then gives either a conclusion or stays at "we do not know". And if the facts say something not PC or unpleasant, this is no valid reason to discount the conclusion.
That way to handle reality is the only known way to arrive at a good approximation of the truth. Yet this approach and way of thinking is completely alien to most people. They seek to have their preconceptions validated and ignore facts that would challenge these. Explains nicely why completely ridiculous ideas get so much people that "believe" them.
The transcript probably made it obvious how very much a fake this whole thing is.
Nice one, new to me. Does also describe why things are so fucked up when you take into account how many lawyers are in politics.
Indeed. Politics is not about being right, it is about sounding right. That, apparently, is a skill that does not need intelligence, just training.
But debating requires better understanding of natural language than ordering pizza. To successfully rebut an argument, you need to understand its logic. I don't doubt that there's room for improvement, but it's a non-trivial step.
You misinterpret what you saw. This demo just shows that "debating" one of a topics that is previously known does not require intelligence. It also casts some real doubt on this type of competition and indicated that the "debates" done there are not actually debates in the sense most people understand that word. This machine can extract logic, but it cannot understand it. It is basically ordering of pizza on a larger scale, but with as many understanding of the nature of the action, i.e. none.
Indeed. It basically is for building cheaper (not smarter) expert systems with an natural language interface. As soon as you go off-script, this technology gets completely lost.