>>47
Suicide would just give highly unreliable results. You could of course perform quantum suicide experiments (if you think MWI is true), but since you are supported by your brain and body, it's likely that the most probable outcome would be just having you survive in unusual ways or merely somehow end up in worlds where you chose not to kill yourself (also 'natural' worlds such as our own might be a lot more numerous than any 'artificial'-like worlds as given in the example in that novel).
Do you think that a digital abstraction created by scanning one's brain and accurattely rebuilding a model of it (and the body, and the environment) would give a conscious substructure, a substrate independent mind(SIM)? If computationalism is true, then it would be as conscious as you are now. If you think it's false, you would die during the scanning process and never attain consciousness.
I don't think this is a trivial question. The only way to know if the abstraction is conscious, is to experience it (be the abstraction). If I didn't acknowledge my own consciousness internally, I would not think humans (or any other being implementing a brain in any way) could be conscious (or possibly, it would reduce to something as trivial as having internal state in a particular reflective manner). I don't believe in continuity of consciousness as an intrinsic property, so I don't treat the procedure as anything different from someone going under anesthesia, assuming that the scanning/conversion process is done as accurately as possible.
The particular experiment described in Permutation City is more or less the same kind as the one that first tests if a computational abstraction would be conscious (the only one that could have any right to claim that would be the abstraction, it cannot be tested in any scientific manner, except of course, in making sure the brain/body abstraction behaves as close as possible as the original brain/body). In that experiment, the substrate independent mind (and its model-of-body/environment + OS) is placed in a self-contained mathematical/computational object whose time evolution can be described entirely by its initial state and future possible states. In that book, the reason for possibility of existence of uncomputed states is explained by first showing that consciousness is not spatially or temporally located, and not represented as anything, except that particular abstraction. A similar (but considerably more rigorous) thought experiment is shown in Marchal's papers (I've linked one before), or if you want it in a simpler form:
http://groups.google.com/group/everything-list/msg/e422b8cef00b3aa6 .
The main difference between 'Dust Theory' (PC) and UDA (see link above) is that 'Dust Theory' depends on our particular physical universe existing (so states could be found anywhere in it), while UDA shows that this requirement is unnecessary (instead a form of Arithmetical Realism (basically CUH) ends up being used after it's shown to be a much simpler alternative). I should probably add that the model used in PC for the structure of their universe is not really ideal and there are some much better choices that could have been used (although probably wouldn't have made as entertaining fiction if there's really no problems to solve; I could elaborate on what I would propose to be used instead, but I think I'd be getting way offtopic by going into that here).
To sum it up, a conscious substructure existing as a computational abstraction (for example running on a real computer/hardware in this world) means computationalism is true. If computationalism is true, CUH is also very likely true (as shown by UDA). If CUH is true, any such self-contained world would also exist and self-aware substructures within it would exist independently of our universe, yet by performing such an experiment, you could have such a substructure which thinks it came from this particular universe (of course, including whatever was in the new universe's initial state). I should also note that there are various potential failure modes (the one described in PC would be highly improbable, but as quantum mechanical-like worlds are likely to be observed, violation of the laws that were set in the initial state are likely to occur, at least as long as one's consciousness is still supported by the physics(Anthropic Principle); there are some ways one could mitigate these risks and end up in a stable world by the way the actual universe's laws are selected, but this is again outside the scope of this post...).
You might also ask yourself why would such a SIM find itself in the particular universe you've coded for? Maybe there's no reason for this and they would find themselves (different copies) in different universes, as long as they support the SIM's existence (Anthropic Principle). On the other hand, there are various ways to minimize the risk of ending up in an universe where you don't want to exist. It may be that our consciousness is so stable because in our current (MWI) world, that the number of copies of oneself increases exponentially, or maybe it has to do with our universe being of low complexity and thus more probable to experience (both views have their proponents).
Either way, if you don't believe computationalism can work, you think mind uploading is equivalent to killing yourself and copies won't be conscious. If you think they can, then there is a good chance the experiment in "Permutation City" would work.