Joel wrote: >What is weird from one perspective is not too weird in another. Some might >thing it's strange, and others might not.
I agree. So what I say is that we must explain why the "world" seems to *remain* normal to us. Suppose you have a theory of elementary particles, and that your theory predicts that if you send a neutron on a proton with enough energy, then you get a virtual mad cows lasting for 0.00134sec. And then you do the experiment, and you get nothing (except the proton and neutron). You agree that you must explain the absence of madcows, I guess. Well, with everything-like type of TOEs, there is a necessity to explain the apparence of lawfull regularities, because those TOEs (like the Everett one) a priori entails to much possible continuations, there is an induction-inflation. In this list, oversimplifying a little bit, there are two sort of approaches with respect to that inflation. 1) The Schmidhuber-Ruhl-Dobrzelewski-... approaches (SRD). (Please don't take such denomination too seriously). The SRD makes abstraction of the first person and does not take into account the first person description. There is some agreement that there are still third person white rabbits, and they hunt them by using some absolute self-sampling assumption (Nick Bostrom SSA, see http://www.analytic.org/) based on universal prior. 2) The other approaches, which take into account the fact that we have not yet solved the mind body problem. In particular if we accept the computationalist thesis, then it is possible to show explicitily that consciousness cannot be associate neither with "physical activity", nor even with any single computation, but only to dense and continuous sheaves of infinite computations. You still have to explain the absence of the third person whabbits (and universal prior could still be useful although my own track of 3-whabbits is more based on Bennett notion of depth), but, you get 1-person whabbits too (and a priori vastly numerous, at least 2^aleph0). To track them you need a refined *relative SSA*, which can be seen as a conditionalisation on your "actual" states. Some are hunting the 3-whabbits, and some are hunting the 3-whabbits + the 1-whabbits. No doubt that that summary is rough, but I wanted to be short. Bruno