X-Message-Number: 23981 Date: Sat, 24 Apr 2004 23:26:38 -0700 From: Mike Perry <> Subject: Computers and Brains again Replying to Thomas Donaldson's #23952 : >In what sense is QM a formal system? Just as with any other phenomena >in the world, we can produce a formal theory which describes how it >works --- so far as our experiments with it go. I think you've answered your own question. It is indeed a formal theory. > However physical >theories should never be confused with reality. It's not that physical >theories aren't very useful; they just aren't the same. I agree. >Furthermore, computers by their nature are supposed to follow Turing's >Theorem. This at least looks like a much stronger requirement than >following a formal system (theory?). They are designed to follow the Turing computational model as far as possible. > Chess follows a formal system >in that its rules can be set out and breaking them can be clearly >verified. Does the game of chess then constitute a computer? Careful >here: it's not enough to say that it can be played on a computer. >So can celestial mechanics, on a large and parallel computer, but >no one claims that celestial mechanics IS a computer. (At least >no one whom I would consider sane). I wouldn't claim chess "is" a computer. >And since this subject began with human beings, I will discuss >humans here too. Yes, human beings like everything else must follow >physical law, chemical laws, biochemical laws, and all the rest. >Do human brains then work like ANY kind of computer? Not any computer we have today, and maybe not as any computer of the future, depending on how the meaning of that term is extended as progress continues. But so what? To me a more important question is whether a sufficiently advanced computational device could, for all intents and purposes, emulate a human brain. Other interesting questions: Would such devices, if they could be made, function better (faster, say) than the human brain? Would they make good replacements for our fragile brain tissue? Is this something we should look forward to? Will it play an important role in (some at least) cryonic resuscitations? And so on. >To make a >computer like a human brain you'd need one which grows new processors >and new connections between both old and new processors, with that >connectivity constituting its memory. I would ask how important in the long run will be the growing of new processes and so on, in terms of the functioning of the system and how well it may, among other things, be able to process information the way a brain does and possibly be substituted for brain tissue someday. The growing of processes, and other self-modifications of hardware, are future possibilities which we might realize in artificial devices. But we may find that the advantages of doing this are not as great as they seem. The same effects may be doable in software, yet offer advantages such as ease of backups, greater versatility, and so on. The fact that a brain now changes its hardware as it does would seem to place it in a class outside of computers, yet again I question how important this distinction really will be in the long run. In particular we don't see brains able to perform any computational task, reliably and reproducibly (or even probabilistically) that computers cannot perform, albeit sometimes more slowly (and other times much faster). >...we, with our brains, >when faced by events which don't follow our formal systems, >abandon them to make other ones which fit better. It's our ability >to do that which provides evidence that we aren't computers, we >just use them. No, Thomas, computers can do this very thing too--abandon one formal system for another and so on. This is where heuristics come into play. This could happen with a robot with an onboard computer, as it confronts and adapts to its environment, and performs various tasks it is motivated to do. It becomes programmed by its own past experience, so that its behavior in effect is unpredictable and resembles that of a living organism. (So far the sophistication of such devices is quite limited compared to natural systems, but how significant is this? A colony of robots, by the way, could be formed that would have its own independent behavior, including adaptations over time, and not simply be "used" by humans.) That is not to say that computers cease to be classical computational devices, whose behavior overall is explained by Turing's theory. But I would offer here that the illusion that the brain is doing something fundamentally beyond the reach of computers comes from the brain's great and inscrutable complexity. Brains, it can be conjectured, operate according to the formal system QM, at least closely enough that they can be taken as explained by that system. I don't have the details, but I think that can be turned into an argument that brains are not in a class apart from computers, albeit the computer is constructed so its inner workings *are* known to us in ways that brains are not--as yet. Endless best to all, Mike Perry Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=23981