[What about Bob?]

I hope everyone will actually READ the entire piece before flame-broiling me.

[What about Bob?]

for your reading pleasure

[garbanzo]

One of the big problems with AI research was the belief that minds were algorithmic computational devices. However our brains are networks of simple "computation" devices if you will. The behavior of neural nets is vastly different from the behavior of processor-based algorithms and the limitations of algorithms don't quite apply.

[Forgiven_Sinner]

I've read "Shadows of the Mind" and this is a very good summary of it. The author is more sympathetic to Penrose's book than most reviewers; most reviewers are staunch materialists and can't stomach Penrose's incontravertable argument against a materialistic mind. Most reviewers (this is about the fourth review of the book I've read, besides the endless discussion it engenders on Usenet's comp.sci.ai.philosophy) get side tracked by Penrose's propositions of a new quantum mechanics, rather than addressing his arguments against a materialistic mind.

[BADROTOFINGER]

Nothing like a good article like this one on FR to make me feel very dumb. Oh well, more books for the already long "to read" list...JFK

[xm177e2]

That is one of the WIDEST posts I've ever seen.

Does everyone know there are more articles in the golden box of text, if you scroll to the left?

[VadeRetro]

I suppose I should be talking to Penrose and not you but I'm not real impressed with this:

Now suppose that there could be a computer program that could perform all the mental feats of which a man is capable. (In fact, such a program must be possible if each of us is in fact a computer.) Given sufficient time to study the structure of that program, a human mathematician (or group of mathematicians) could construct a "Godel proposition" for it, namely a proposition that could not be proven by the program but that was nevertheless true, and-here is the crux of the matter-which could be seen to be true by the human mathematician using a form of reasoning not allowed for in the program. But this is a contradiction, since this hypothetical program was supposed to be able to do anything that the human mind can do.

The contradiction was introduced just before in the thought experiment's gimmee test condition:

. . . but that was nevertheless true, and-here is the crux of the matter-which could be seen to be true by the human mathematician using a form of reasoning not allowed for in the program.

Sneak it in and "Ta-dah" on it in the next sentence! If the program is a perfect human emulator, you can't disallow it any form of logic humans can do. There may be a Godel proposition for the program, but it's a Godel proposition for the human too so there's no distinction.

Maybe it's the article-writer's fault. I tried to read Penrose's The Emperor's New Mind and gave it up about 2/3 through. He's like a professor that gets lazy about making sure the class is still with him and starts to just drone on while the students, lost, doodle in their notebooks.

I don't buy that the mind is not a machine operating under the laws of physics. Maybe I just need to see the proof written up a little better.

[RussP]

Very interesting stuff. But I don't think you need to get quite so technical to prove that science can never answer certain very fundamental questions.

Ask yourself this simple question: can science explain, or will it ever be able to explain, why you are you? This may seem silly at first, but it's actually a very important question.

When you were conceived, why did you end up your particular physical body - and not another? I don't see how science will ever be able to explain that one. Do you?

By the way, if our minds (and everything else) are nothing but matter, then what is truth? Is it matter too? If you think it is something other than matter, than you are not a strict materialist. And if you think that truth is nothing but matter, then you shouldn't care about it any more than you care about a pile of dirt, should you?

[sourcery]

The relevance of all this to computers is that all computers involve- indeed are-systems for the mechanical manipulation of strings of symbols (or "bits") carried out according to mechanical recipes called "programs" or "algorithms." Now suppose that there could be a computer program that could perform all the mental feats of which a man is capable. (In fact, such a program must be possible if each of us is in fact a computer.) Given sufficient time to study the structure of that program, a human mathematician (or group of mathematicians) could construct a "Godel proposition" for it, namely a proposition that could not be proven by the program but that was nevertheless true, and-here is the crux of the matter-which could be seen to be true by the human mathematician using a form of reasoning not allowed for in the program. But this is a contradiction, since this hypothetical program was supposed to be able to do anything that the human mind can do.

Penrose's argument is based on several logical fallacies:

• It assumes, without any basis, that any purely material brain must necessarily operate as a Turing machine. In fact, there is no reason whatsoever to believe that this must be true. The article claims that Penrose has proven that all physical processes must be "computational." It's true that some scientists believe that all physical processes are computational, but it's not true that consensus exists that anyone has proven this. But in any case, the unstated implication that all physical processes are equivalent to Turing machines is known to be false.
• It relies on an inconsistent usage of terms such as "physical" and "material." Sometimes these terms are used to mean "anything which is real." Other times, they are used as antonyms of "spiritual" and/or "abstract." Penrose's proof is based on slyly using these terms with first one meaning, and then the other. Unfortunately for the validity of Penrose's argument, an argument that uses terms with inconsistent semantics is invalid on its face.
• Penrose's argument involves an operation not permitted by his initial assumptions: he requires that human mathematicians construct a proposition that the human mind cannot prove to be true (according to the initial assumptions), but that these human mathematicians nevertheless know to be true. But if their human minds are not capable of proving such a proposition, then it is impossible for them to know whether or not the proposition is true. Conversely, if these human mathematicians do know these propositions to be true, then they must be able to formulate a proof--and if they can do that, then the human mind is in fact capable of proving the proposition. So yes, Penrose has demonstrated a logical contradiction--but instead of proving his argument, it invalidates it.
• Any entity that could prove propositions that are outside human competence to prove (true or false, makes no difference) would simply have minds whose mathematical power exceeded that of a human mind. Godel's proof permits such propositions to be proven by going "outside the system." There is no reason that material minds cannot exist whose principles of operation transcend those of humans, just as ours transcend those of ants. Our ability to prove what ants cannot no more implies that our minds are immaterial than does the fact that the aliens from the next galaxy can prove what we cannot.

[VadeRetro]

But his [Penrose's] materialist assumptions have painted him into a very tight corner.

. . .

Penrose is all the more effective in overthrowing materialism because that is not his aim.

. . .

There are several misconceptions here [in Penrose's statements on "mysticism"] . . .

We're seeing the usual hijacking of a more-or-less sober scientist by an "anti-materialist" whose axe grinds loudly throughout the article. Barr loves Penrose for giving him some good ammo against that evil, materialism. Then he gets mad at him for not going far enough, for not rejecting materialism himself.

Yes, Barr is a Physicist at U. Delaware. He also writes articles about the anthropic principle proving the universe was made for us. That doesn't wash with me either but it's another story.

[Aquinasfan]

The idea of "thinking machines" is a category error. End of story.

[dubyagee]

In this view, religion has been fighting a long rear-guard action against the advance of knowledge, taking refuge in the unknown and the obscure by positing a "God of the gaps," and, as the gaps in our rational explanation of the universe disappear, God will be driven out. This is indeed one of the main motivations for a certain kind of scientist who supposes that when the job of Science is done there will be no room left for the "superstition" of religious belief.

I personally am not worshiping "a God of the Gaps". I think the fact that there exist laws in which this universe operates, is incontrovertible PROOF that God exists. Where there exist laws, there must be something "creating" those laws. My problem is not with the LAWS of science, or with scientists in general, but with those who choose humanism or objectivism as religion, and then deny that it is such. To back up their religion, they feel they must deny God, and in denying God, they feel they must belittle those who choose to believe God. Therefore evolution is the basis for their argument. All those who believe otherwise are intellectually challenged. Or so they believe.

[WRhine]

Religion supplies irrational explanations where rational ones are lacking; as lightning, for example, is still thought by primitive people to be the raging of the gods.

Well I got this far into the article. I wish I had more time to read it. It was interesting in some respects despite its noticeable defensive tone. Now as for the above, I'd like to ask the Creationists lurking here to explain how using Religion to supply irrational explanations for the very real and documented evidence supporting Evolution is any different than primitive people using their religious beliefs to explain away lightning as the raging of the gods? Or do most Creationists still believe that lightning is the raging of gods? And if not, what does that tell you?

[the808bass]

I hope everyone will actually READ the entire piece before flame-broiling me.

I've often thought that shopping at pseudo-trendy stores in malls and a belief in the Almighty are mutually exclusive. I guess this article proves it.

Someone had to say it...

[steve-b]

Now suppose that there could be a computer program that could perform all the mental feats of which a man is capable. (In fact, such a program must be possible if each of us is in fact a computer.) Given sufficient time to study the structure of that program, a human mathematician (or group of mathematicians) could construct a "Godel proposition" for it, namely a proposition that could not be proven by the program but that was nevertheless true, and - here is the crux of the matter - which could be seen to be true by the human mathematician using a form of reasoning not allowed for in the program.

Contradictory premeses lead to a contradictory result. Well, duh.

[Junior]

Bump.

[Phaedrus]

Bump for an excellent article -- alternate title: "Materialism on the Run".

[Nebullis]

Penrose carries some fear about intelligent machines into a call for a new science, a science not computationally based, and thereby opens the door to the introduction of a way in which we may actually make intelligent machines. The clamor for a new paradigm is a loud one in the AI community; not much profound has been accomplished in the last 10 years.

On the other side, it is a prediction of materialism that human beings can have no free will. The behavior of purely physical systems as it has ever been described by science involves only two possibilities: the regularity of deterministic laws or the randomness of stochastic laws. There is no room for the tertium quid that is free will. But this prediction of materialism is falsified by the data of our own experience: we actually exercise free will, and thereby know it to be other than either determined or random.

First, it is not a prediction of materialism that human beings can have no free will. (Where does Barr get this stuff?) The behavior of physical systems is either deterministic or random, and free will does not fit either?

But what is free will? The ability to rationally choose between alternatives? And why should we expect this to be described by physical laws governing the atom? Pressure is not described by the behavior of molecules in a gas. It is the emergent property of a collection of molecules. Free will, or the ability to rationally choose between alternatives, is a much higher level emergent function. But the dog has it. And the fuitfly does. And the nematode?

[Nebullis]

Incidentally, I don't think that we'll be producing the human mind in AI. I suspect that wouldn't be very useful. Or scary. The hope is that we can improve, so this already requires something novel. Perhaps the novelty that Penrose is looking for (not God).

Machines, as computational powerhouses, already far surpass the human brain. The limitations of the mind may well be augmented by machines. Maybe even machines which can distill the complexity of biology and cosmology into 3 or 4 dimensions for us.

[jammer]

It was a pleasure. Thanks, Bob. Into the bookmarks.

[jennyp]

So, let's assume Penrose is right and the brain can't be a Turing machine. I'm wondering (not a computer theory expert) why should this be a problem? The brain is a network of digital/analog components. The neurons aren't strictly digital. Each neuron integrates the pulse trains coming into it, where the frequency of the pulses represents a positive or negative amount that adds to or subtracts from the electrical level.

Has anyone ever claimed or proven that an analog computer (or analog/digital hybrid) would also be a Turing machine?