Excellent jessem, thanks. This line from the abstract of the first paper you cite pretty much summarises the changed understanding of MWI I was getting at:
Measurement-type interactions lead, not to many worlds but, rather, to many local copies of experimental systems and the observers who measure their properties. On Thursday, January 23, 2014 3:24:14 AM UTC+11, jessem wrote: > > > > On Tue, Jan 21, 2014 at 10:34 PM, meekerdb <[email protected]<javascript:> > > wrote: > >> On 1/21/2014 4:50 PM, LizR wrote: >> >>> It seems to me that differentiation is local, and spreads slowly, and >>> that there is always going to be some remerging (but only in proportion to >>> the chances of entropy reversing). The an atom starts in a superposition of >>> decayed and non-decayed. Now a cat is in a superposition of alive and dead. >>> Now an experimenter is in a superposition of having seen an alive and dead >>> cat... now everyone who reads "Nature" is in a superposition ... but none >>> of this affects Jupiter for a long time, >>> >> >> Does it? Suppose there's an electron on Jupiter that was entangled in a >> singlet state with an electron on Earth and the electron on Earth just got >> it's spin measured? MWI may be able to model this with a local hidden >> variable, but in THIS world it looks like FTL influence - and it can go a >> lot further than Jupiter, e.g. the CMB. > > > There's no need even for hidden variables to explain this in a MWI > context, as I understand it. Here's a pair of technical papers on the > subject by David Deutsch: > > http://arxiv.org/abs/quant-ph/9906007v2 > http://arxiv.org/abs/1109.6223 > > And a few more papers on locality in (nonrelativistic) quantum field > theory by another many-worlds advocate, Mark Rubin (p. 2 of the first paper > below has a good summary of other work by MWI advocates on the subject of > how locality is preserved): > > http://arxiv.org/abs/quant-ph/0103079v2 > http://arxiv.org/abs/quant-ph/0204024 > http://arxiv.org/abs/0909.2673 > > I think the basic conceptual explanation is something like this: in your > example of the entangled electrons on Earth and Jupiter, when an > experimenter on Earth measures an electron, the experimenter locally splits > into multiple versions who may see different results from one another, and > likewise with the experimenter on Jupiter. And there is no need for the > universe to decide which version on Earth will be part of the same "world" > as which version on Jupiter until there has actually been time for a > physical message (moving at the speed of light) to pass from one to the > other. > > I can illustrate this with a simple toy model. One of the various Bell > inequalities says that if experimenters at each location can measure spin > at three different detector angles, and on every trial where they choose > the same detector angle they always find opposite spins, then on the subset > of trials where they choose two different detector angles, the probability > they get opposite results must be greater than or equal to 1/3. But in QM > it's possible that they do always get opposite results with the same > detector angle, but the probability they get opposite results when they > choose different angles is only 1/4, which violates this Bell inequality. > But now let's suppose we want to simulate this using a classical computer > simulation, using AI experimenters running on computers on both Earth and > Jupiter (call the AI on Earth "Ellen", and the AI on Jupiter "Jim"). > Suppose each AI uses a pseudorandom algorithm to decide which choice of the > three detector angles they decide to use on each trial. Unbeknownst to the > AIs, though, each time they make a simulated measurement, the program > creates 8 different copies of that AI, 4 of which get the result "spin-up" > for the measurement axis they chose on that trial, and 4 of which get the > result "spin-down". We can assign the copies numbers to differentiate > them--so Ellen #1 got spin-up, as did Ellen #2-4, and Ellen #5-8 got > spin-down. Likewise Jim #1-4 got spin-up, and #5-8 got spin-down. > > After the Ellen on Earth gets her measurement result, she wants to > communicate it with the Jim on Jupiter, so she sends a message which > travels to Jim at the speed of light, telling him both her choice of > detector angle and whether she got spin-up or spin-down at that angle. But > unbeknownst to Ellen and Jim there are actually 8 different versions of > each of them, so from our point of view outside the simulation, we see that > what actually gets sent is a bundle of 8 parallel messages, and when they > arrive at Jupiter, the simulation has some algorithm to assign one of the 8 > parallel messages to each of the 8 parallel versions of Jim. The key is > that the simulation's algorithm can work in such a way that over the course > of many trials, each copy observes statistics that violate Bell's > inequality, even though this is a purely classical simulation (because > Bell's proof assumes a unique measurement result at each location, which is > violated here by all the copies). On trials where they both chose the same > detector angle, the simulation matches up the messages like this: > > Jim #1 (spin-up) gets the message from Ellen #5 (spin-down) > Jim #2 (spin-up) gets the message from Ellen #6 (spin-down) > Jim #3 (spin-up) gets the message from Ellen #7 (spin-down) > Jim #4 (spin-up) gets the message from Ellen #8 (spin-down) > Jim #5 (spin-down) gets the message from Ellen #1 (spin-up) > Jim #6 (spin-down) gets the message from Ellen #2 (spin-up) > Jim #7 (spin-down) gets the message from Ellen #3 (spin-up) > Jim #8 (spin-down) gets the message from Ellen #4 (spin-up) > > The above matching ensures that every single copy of Jim gets a message > from Ellen saying she got the opposite spin result. But on the trials where > they chose different detector angles, the simulation matches up messages > like this: > > Jim #1 (spin-up) gets the message from Ellen #1 (spin-up) > Jim #2 (spin-up) gets the message from Ellen #2 (spin-up) > Jim #3 (spin-up) gets the message from Ellen #3 (spin-up) > Jim #4 (spin-up) gets the message from Ellen #5 (spin-down) > Jim #5 (spin-down) gets the message from Ellen #4 (spin-up) > Jim #6 (spin-down) gets the message from Ellen #6 (spin-down) > Jim #7 (spin-down) gets the message from Ellen #7 (spin-down) > Jim #8 (spin-down) gets the message from Ellen #8 (spin-down) > > This matching ensures that 3/4 of the copies of Jim will see that Ellen > got the same spin as himself, while only 1/4 of the copies of Jim will see > that Ellen got the opposite spin. These are the statistics that would be > seen in QM, the ones that violate the Bell inequality. > > Jesse > > -- 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 post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/everything-list. For more options, visit https://groups.google.com/groups/opt_out.
