On Fri, Jun 21, 2019 at 1:19 PM 'Brent Meeker' via Everything List < [email protected]> wrote:
> On 6/20/2019 5:56 PM, Bruce Kellett wrote: > > From: Bruno Marchal <[email protected]> > > > I don’t think your refutation of step 3 has been understood by anyone. > > If someone else want to argue that there is no indeterminacy in the self > duplication experience, he is welcome. > > > I think that some might challenge your interpretation of this > indeterminacy. This might not be exactly JC's objection to step 3, but, to > my mind, it is a serious difficulty in its own right. > > This comes from a recent podcast of a conversation between Sean Carroll > and David Albert: > > https://www.youtube.com/watch?v=AglOFx6eySE > > This is a long discussion, and the relevant parts of Albert's objections > to MWI and self-locating uncertainty come towards the end. > > The essence of Albert's point is that in the duplication case, you ask > "What is the probability that you will find yourself in Moscow (resp. > Washington)?" Putting aside objections to the non-specificity of the > pronoun 'you', I think your answer is that the probabilities are 0.5 for > either city. Albert points out that to reach this conclusion, you use some > principle of indifference, or point to some symmetry between the possible > outcomes. Using this symmetry, you claim that the probabilities must be > equal, hence 0.5 for each city. Now, says Albert, there is another solution > that also respects all the symmetries involved, viz., "I have not idea what > the probability is." > > You can then easily argue that this is a better solution. Because the > probability 0.5 is not written in the physics of the situation -- it comes > entirely from the classical principle of indifference. So Albert asks how > you are going to verify this probability experimentally -- as a large N > limit, or something similar. So you repeat the duplication N times on your > participants. i.e. after the original duplication you transport the > subjects back to Helsinki and repeat the duplication to Washington and > Moscow. You end up with 2^N copies, each of which has a record of the N > cities they found themselves in after each duplication. You now ask each of > them their best estimate of the probabilities for W or M on each > duplication. Of course, you then get all possible answers, from 1/N for M > to 1/N for W. Since, withprobaility one, the will always be someone who > found himself in M each time, and similarly, someone who found himself in W > each time. Plus all other 2^ possible combinations of results. > > > But most participants will say they were in Washington approximately N/2 > times and Moscow N/2 times, in accordance with a binomial distribution. > But I am not "most participants". I am just me, only one of me. I could easily be the guy who sees 100% Moscow. I am the one you have to convince, not those who saw different things. Bruce > > Brent > > So who has the correct estimate of the probability for W or M on each > duplication? Clearly, the guy who says "I have no idea" has a better grasp > of the situation than the guy who confidently claims, "The probability for > M is 0.5, and similarly for W." These probabilities are not written in > stone, and any attempt at an empirical determination of the probability > will necessarily yield all possible results. > > This is the core problem with understanding the origin of probabilities in > MWI -- self-locating uncertainty is not good enough when all outcomes occur > with probability one. > > Bruce > > -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/everything-list/CAFxXSLSOAEVXG3nSHjag_Su0KdkbKojHnKYR5tamgc6yt2eQyQ%40mail.gmail.com.
