On 9 January 2005 Alastair Malcolm wrote:
This is a fascinating discussion list, full of stimulating ideas and theories, but I would be interested to know what people *actually* believe on the subject of many/all worlds - what one would bet one's house or life on, given that one were forced to choose some such bet.
I believe it is logically necessary that all possible worlds exist, based on a number of ideas that have been discussed many times on this list:
1. Every possible world can be simulated by a computer program.
2. It is not, in general, possible to distinguish between a simulated world and a "real" world.
2a. If there is no empirical or logical way to distinguish with certainty between a real and a simulated world, one may as well say there is no essential difference between them [interesting, but not really necessary for the rest of this argument].
3. Consider a computer running a simulation complete with conscious beings. This particular computer was designed by a now extinct civilization, and although the hardware still appears to be working, the compiler, instruction manual and computer language documentation have all been lost.
3a. The result of (3) is that the simulated world continues to run on the computer, even though there can be no communication between it and us in the "real" world where the computer exists physically. (Idea for a story: maybe that's why the gods used to talk to us in ancient times but no longer do!)
4. A computer need not be a box that runs Windows or Linux. Conceivably, a computer could consist of the idle passage of time, or the set of natural numbers, operated on by some hugely complex look-up table. In Greg Egan's 1994 novel "Permutation City", it is pointed out that a simulated being's experiences are the same if the computation is run backwards, forwards, chopped up into individual pieces and randomly dispersed throughout the world-wide network; the computation somehow assembles itself out of "dust" - out of omnipresent, apparently randomly distributed ones and zeroes.
4a. In other words, no hardware, whether physical or simulated (if these are different things) is necessary for the implementation of a computation. Every possible computation is implemented out there in the realm of pure mathematics, so every possible world necessarily exists.
Now, the above argument is sometimes taken as being self-evidently absurd. Hilary Putnam and John Searle have actually used it in this way to attack strong A.I. theories. The objection is that the effort and information needed to construct the look-up table in (4) is at least as great as that which would be needed to construct and program a computer in the conventional way. An analogy can be made with a block of marble: in a sense, it does contain every possible statue, but this is not any help to the sculptor. But all this means is that we cannot, in this world, communicate with or make any use of the type of computer described in (3) or (4). The other worlds may exist, but we can never know this directly.
As for the "failure of induction" if all possible worlds exist, I prefer to simply bypass the problem. I predict that in the next few moments the world will most likely continue to behave as it always has in the past... Here I am a few moments later, and I am completely, horribly wrong. A zillion versions of me in other worlds are dying or losing consciousness as they watch fire-breathing dragons materialise out of nothing. So what? Those versions are not continuing to type to the end of this paragraph, while this one-in-a-zillion version manifestly is, and will continue to live life holding the delusional belief that the laws of physics will remain constant.
Searching for that dream home? Try http://ninemsn.realestate.com.au for all your property needs.