X-Message-Number: 8240 From: (Thomas Donaldson) Subject: Re: CryoNet #8237 - #8239 Date: Sat, 24 May 1997 18:09:19 -0700 (PDT) To Mr. Metzgar: Perhaps I should give your comments the brief attention you seem to have given mine. If you have been following this issue on Cryonet, you will know that I personally have not stated that we cannot build intelligent, conscious devices, but that the architecture of most current computers would make that extremely difficult if not impossible. FURTHERMORE, if you insist that a study of consciousness must necessarily include all possible forms of consciousness, by animals, by machines, by some creatures inhabiting the moons of a planet of Vega, YOU are guilty of religious thinking. If we want to understand something, one good way to do so is to try first to understand it in one special case. Since so far no one has actually produced a conscious computer, and we have no access to the inhabitants of the moons of Vega, we are left with one kind of animal to start our study with: humans and primates generally. To do this is not to claim that other kinds of creature or machine cannot be conscious. It is a means to start our empirical investigation. Did I ever "define consciousness"? Not in the global sense you seem to want, and which I think would be foolish without much more information. And with this notion, I suggested an experiment: that awareness requires activation of a particular brain area. This experiment would actually be expensive to do, but remains doable in theory. So if you remain interested in the consciousness of computers, here is a little program for you. First we work out the details of how OUR HUMAN brains work in terms of awareness-consciousness. If we understood these details, it should be easy to make a device --- whether we want to call it a computer or not is not clear --- also capable of the same awareness-consciousness as our own. Not only that, but it will give us strong hints as to where to look for this trait in other animals: do they retain this brain structure, and when they show all the signs of responding to the world around them, is it activated? We would even be able to work out just what this structure does: as I have suggested before, it probably is not a "seat" of consciousness but instead a circuit element through which a lot of processing must pass if the animal/man is conscious. As for Searle, I do not want my arguments to stand or fall on the basis of what Searle said, or Turing said, or Descartes said, or whatever. I want my arguments to depend on what *** I *** said. The Turing Test has what I consider a major flaw. This does not mean that I think computer awareness is impossible, it means that the Turing Test has a major flaw. No more and no less. Depending on just what Searle thought would qualify as a computer -- as opposed to qualifying as a human being --- his claim may or may not have made sense. But that is a side issue. Not only that, but as I said clearly in the previous paragraph, once we understand "consciousness" in primates (including humans) we will also know how to make devices which are conscious. It is my knowledge of neuroscience, such as it is, that makes me doubt that these devices would qualify as "computers", but that statement should be evaluated on its own merits without regard to what some other guy has said or not said. As for the creatures living on the moons of planets of Vegas, I will give here a speculation about how consciousness MAY work which MAY apply even to them. (If you don't like speculation I wonder why you're here). If their multiple or single brains are so wired and work that they have one single sequential process among many parallel processes, and that sequential process acts to choose between different goals and desires on the basis of results from all the other parallel processes, then I would say that they are conscious. This requires that they have not just knowledge but desires, and thus have some circuits by which their sequential process can become informed of these desires. I make no commitment about how fast or slow they may be, nor about just what these desires may be. An intelligent plant will no doubt seek sunlight and a place containing its requirements of minerals, etc. So long as this seeking requires a choice, and the plant is set up able to make such a choice, then it has awareness. So far as I know, no earthly plant does this. It might be hard to see, though, because it would happen much more slowly than our own brains work. Speculations cannot be disproven as such. No matter how many plants you show me that lack the required circuits, I'm saying that some MIGHT have this feature. But it is a feature, and by understanding just how the plant works we might find out how to test for it. Without that understanding, it remains only a speculation. By the way, is speculation religious or scientific in your scheme of things? Finally I will point out that various people have come up with many other ideas about consciousness. IF my idea proves to be correct, we may see the existence of this special sequential subsystem, and the presence of desires, as providing a DEFINITION of consciousness. If not, not. Without allowing any proposed theory of consciousness to exist without first giving a definition, you make the problem quite unsolvable even on your own terms. It's simply not the best way to proceed: though to some it may sound paradoxical, I cannot get any good understanding of consciousness (or in physics, of mass) until I have at least an approximate theory of how consciousness works (or mass works). The two go together and cannot be separated. To Olaf Henny: Here is the essential point I am making. We are now at time 24 May 1997. I am sure that if someone worked fast enough they could produce a complete list of all known diseases and medical problems. And sure, everyone one of these will eventually (don't know when, and without understanding of the problem I cannot predict when) become trivially curable. Yet that very much does NOT imply that we will then be free of all diseases and medical problems. This is very easy to see if we simply consider past history. What has happened, along with advances in our medicine, has been a BIG INCREASE in the number of diseases and medical problems we can distinguish from one another. Furthermore, many of these problems simply did not exist until most people lived long enough to develop them. Sure, few people in developed countries now die of typhus, but abolition of typhus has not made us free from problems. This history should be thought of as a counterexample. And because it is history, it's a strong one. No matter what science and technology we develop, nano- or other, things will still go wrong. Most likely, they will be things going wrong with the very technologies themselves. In going wrong, they will produce medical problems that could not possibly have been on any list put together on 24 May 1997. (I say this because I do think a day will come when "natural" threats, such as smallpox or Ebola, will be easily fixable --- but we will still live in a world in which things sometimes go wrong). THERE IS NO TECHNOLOGY WHICH CANNOT SOMETIMES GO WRONG. There will always be a bug there somewhere. And as we become more skilled, the bugs become fewer (and harder to find or deal with) but they never go away. That is why I think that some form of cryonics will persist for a very long time --- so long that we may as well accept it as indefinite. To claim otherwise is basically to say that someday we will design our machines and our society to be perfect. That is out of the question. Long long life, Thomas Donaldson Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=8240