X-Message-Number: 7952 Date: Wed, 26 Mar 97 12:21:26 From: Mike Perry <> Subject: Re: CryoNet #7941 - #7944 Bob Ettinger wrote > It is simply a fact that NONE > of the usual intuitive survival criteria, nor any combination, can be shown > to be EITHER necessary or sufficient--not continuity of matter, not identity > of matter, not continuity of pattern, not identity of pattern. It is not > possible, and in my opinion not psychologically necessary, to form a firm > opinion before we have the necessary knowledge of biology and physics. So, > once more, we just use common sense, freeze our assets, and hope for the > best, while continuing to investigate. > "Identity of pattern" would, to my satisfaction, be sufficient (not necessary, however, inasmuch as you could be said to survive in a continuer, which is not identical to you). This is a central point of the book I am writing, and I'd be interested in any comments as to why, if what I mean is reasonably understood, it would NOT be sufficient. The objections, as far as I can see, all amount to forms of vitalism or a non-reductionist "further fact" about what constitutes our identity. At any rate, I second Bob's advice to "freeze our assets and hope for the best, while continuing to investigate." Thomas Donaldson wrote > Now look here. Are you really saying that everything is > symbolic? Not representable by symbols, but actually symbols themselves? I am inclined to think that a universal language could be developed to describe "everything." It would describe the quantum state in an economical way. (Indeed, something like this probably already exists.) The quantum state of a bounded system has a finite description. Would I say that such a language "is" reality? I'll say that I think it would be isomorphic, in a mathematical sense, and probably in a rather strong sense (since more than one isomorphism is possible between mathematical objects, and we then must ask whether we are dealing with the "right" isomorphism). If such a language did exist, it would make it clearer how to emulate processes in the real world--it would be straightforward to map a process in a computer that used this language into a "real" process, and vice versa. The possibility of doing this, in principle, seems strongly suggested by arguments such as the finite state machine argument of Tipler, despite some complications with that idea (see below). > I doubt that completely, unless perhaps you have decided to redefine > the world "symbol" so it includes everything... in which case you are > saying nothing at all. Who attaches meaning to these symbols, anyway? > Without that, you have no symbols. > I haven't "decided to redefine the [word] 'symbol'." I am talking about a language over a finite alphabet--each letter being a "symbol"--each "sentence" being a finite string of such symbols. The alphabet could just as well contain 2 symbols only, say O and X. As for "who attaches meaning"--consider the following string of symbols: OXOXXOXXXOXXXXOXXXXXOXXXXXXOXXXXXXX Now consider this string: OXXOXXXOXXXXXOXXXXXXXOXXXXXXXXXXX I haven't "attached any meaning" to either one, yet I bet you could say something "meaningful" about both--they are not just random noise. More generally, I think that strings of symbols could encode mathematical properties that should be decipherable to intelligent aliens who have no knowledge of our languages or civilization, or what we are like physically. Mathematics would make a powerful starting point for conveying ideas in general, and I think this could be done too. We might then say that a message in such a language would "have a meaning" we didn't have to "attach" in the sense that aliens could pick up on it without our instructing them, other than through the message itself--and other messages in the language. Reality is like that too. If we could reduce its "messages" to pure digital form, and think of them that way, it wouldn't preclude our finding "meaning" in them. > As to whether or not we are finite state machines, I remain dubious. One > issue coming directly from quantum mechanics is that not ALL measurements > (in theory) have a quantum character. For instance, momentum is a vector, > and its value as a vector -- a direction in space --- is not quantized, > though the relation between position and momemtum is. That alone raises > questions about whether quantum mechanics says we are finite state > machines. In practical terms, our components undergo constant renewal, > so that their quantum states will constantly change. This means to me > that any attempt to really list the finite number of states we are in will > fail --- not because of its number, but because it changes all the time. > Sure, we can make generalizations, but then we will lose that finite > character: an infinite set of molecules can be in the right conformation > to produce ME or YOU. > There is a complication with the finite state machine model of real world phenomena, in that such machines must always be "embedded" in their surroundings. (An FSM, however, can be interconnected to other FSMs, through input-output relations.) An infinite universe would not be a finite state machine, nor an immortal person. But the FSM model, even taking account of the (hopefully) infinite character of reality as a whole, does at least strongly suggest that we are digital processes, and thus, exactly emulable in a purely symbol-manipulating device (which must also have appropriate links with the outside world to model the embedding). But this seems contrary to the view you express: > > One major distinction between (say) Crays and human beings comes down to > the issue of whether they operate with symbols or work in the world without > using symbols (at root). I am saying that this distinction is key. > Can you "work in the world without using symbols (at root)"? The message from quantum mechanics, whatever the character of some measurements, etc., seems to be no. This is because the quantum state is held to provide the complete description of a system--or, as Tipler says, to *be* the system. The description in turn is finite (again, for the bounded case), and the way it changes over time is describable by an FSM. The states change rapidly, as you say, but we could, in principle, enumerate all the *possible* states a system could be in for a long stretch of time. We would then know that the system is popping back and forth between various of these states according to known rules--even if, in practice, ascertaining exactly which states at which times may be impossible. Again, the practical difficulties don't seem to stand in the way of this being a digital world. Mike Perry http://www.alcor.org Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=7952