Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.2 9/18/84; site calgary.UUCP Path: utzoo!utcsri!ubc-vision!alberta!calgary!cleary From: cleary@calgary.UUCP (John Cleary) Newsgroups: net.philosophy,net.math Subject: Re: Sc--nce Attack (really on minds and computers) Message-ID: <475@calgary.UUCP> Date: Sat, 26-Oct-85 00:32:27 EDT Article-I.D.: calgary.475 Posted: Sat Oct 26 00:32:27 1985 Date-Received: Sat, 26-Oct-85 05:39:59 EDT References: <299@umich.UUCP> <10699@ucbvax.ARPA> <10700@ucbvax.ARPA> <10702@ucbvax.ARPA> <1925@pyuxd.UUCP> Organization: University of Calgary, Calgary, Alberta Lines: 121 > > [Yes, this is exactly the point. Exhibit the Turing machine that > > is claimed to be equivalent to the human mind, and the human mind > > can reason about the system in ways impossible within the system. > > Thus we contradict the assumption that the machine was equivalent > > to the mind.] This is a very crucial point in this discussion I think. This is only true IF we assume that the human mind that is doing the reasoning is not itself part of the Turing machine being exhibited. The problem is that the physical boundary about a human is most unclear. The wiggling of an electron on Alpha Centauri might via changes in gravitation affect the firing of one of my neurons and so alter my behaviour. From this (extreme) example we have to include the whole universe in the description of the human. That is anything which can affect us (and so observable by us) must be included in a complete description of our behaviour. The set of all things observable by us (or potentially observable by us) can validly be called the whole universe. Unfortunately the whole universe includes all entities that can observe us and hence reason about us (remember Heisenberg, if it can observe you then it can affect you). The interesting thing about digital computers is that we confuse two things, the actual physical machine and its abstract description. The physical machine just like a human needs the whole universe included in it to describe it. The abstraction (what is described in the manuals) is an approxiamtion only. It is proably unclear from the abstract description what happens when a high energy gamma ray passes through the CPU chip. So I agree with those who say a digital computer AS DESCRIBED BY A FORMAL SYSTEM cannot have the same status as a human. However there is no reason we know of at the moment why a physical system cannot, indeed as the description of the physical computer includes the whole universe and the humans in it, it already has the same status as the human. This then raises some fascinating questions: 1) Church's thesis that all computers are equivalent to a Turing machine. This is actually a PHYSICAL law (like law of gravitation) potentially subject to a physical experiment. It is conceivable for example that some of the pecualiar effects of quantum mechanics could allow calculations faster than any possible Turing machine. 2) Is the entire universe a Turing machine? 3) Is it conceivable that any thing part of the universe could verify or refute 2)? I am also struck by the similarity of the conclusions of some philosophers from the Eastern tradition that we are all intimately connected with the whole universe. > > I originally asked whether anyone disputed my claim that the human > > mind is not equivalent to a turing machine. After all the negative > > response, I would like to change my question to: > > > > *IS THERE ANYONE THAT AGREES WITH ME THAT THE HUMAN MIND IS PROVABLY > > NOT EQUIVALENT TO A TURING MACHINE?* See above. I think this is a question for the physicists, and potentially subject to physicl experiment. > > "OK, but could a digital computer think?" > > If by "digital computer" we mean anything at all that has a level > of description where it can be correctly described as the instantiation > of a computer program, then again the answer is, of course, yes, since > we are the instantiations of any number of computer programs, and we > can think. No I disagree, here he talks about the abstract machine. > > "But could something think, understand, and so on *solely* > in virtue of being a computer with the right sort of program? > Could instantiating a program, the right program of course, > by itself be a sufficient condition of understanding?" > > This I think is the right question to ask, though it is usually > confused with one of the earlier questions, and the answer to it is no. > > "Why not?" > > Because the formal symbol manipulations themselves don't have > any intentionality... I agree. > ... If and when such machines come about, their causal powers will > derive not from the substances they are made of, *but* *from* *their* *design* > *and* *the* *programs* *that* *run* *in* *them*. [ITALICS MINE] And the way > we will know they have those causal powers is by talking by them and listening > carefully to what they they have to say." This is a fascinating argument, incorrect I think. Certainly in humans much of their abilities come from there experience of the world, learning adaptation. That is much of their state and behaviour is a result of their experience not their genes. I suspect any really interesting computer will be similar. Much of its behaviour will be a result not of its original programming but of its subsequent experience of the world. Unfortunatly again to describe the machines that result we must describe not only their original programming but all their later possible experiences. But they can potentially be affected by anything in the universe. The problem with the current state of computing, robotics and AI is that most computers have little or no interaction with the real world. They have no bodies. Hence they can to a very good approximatin be described by some formal system. Thus many people have a gut feeling that computers are fundamentally different from humans. In their guise as formal systems I think this is indeed true. I think there is also a practical lesson for AI here. To get really interesting behaviour we need open machines which get a lot of experience of the real world. Unfortunately we arent going to be able to formalize or predict the result. But it will be interesting. Sorry about the length of this, but the question seemed too fascinating to let alone. John G. Cleary, Dept. Computer Science, The University of Calgary, 2500 University Dr., N.W. Calgary, Alberta, CANADA T2N 1N4. Ph. (403)220-6087 Usenet: ...{ubc-vision,ihnp4}!alberta!calgary!cleary ...nrl-css!calgary!cleary CRNET (Canadian Research Net): cleary@calgary ARPA: cleary.calgary.ubc@csnet-relay