On Friday, December 12, 2014, meekerdb <[email protected]> wrote: > On 12/11/2014 5:49 PM, Stathis Papaioannou wrote: > >> On 12 December 2014 at 12:22, Jason Resch <[email protected]> wrote: >> >>> >>> On Thu, Dec 11, 2014 at 3:10 PM, LizR <[email protected]> wrote: >>> >>>> On 11 December 2014 at 18:59, Stathis Papaioannou <[email protected]> >>>> wrote: >>>> >>>>> >>>>> On Thursday, December 11, 2014, LizR <[email protected]> wrote: >>>>> >>>>>> Maybe it's a delayed choice experiment and retroactively collapses the >>>>>> wave function, so your choice actually does determine the contents of >>>>>> the >>>>>> boxes. >>>>>> >>>>>> (Just a thought...maybe the second box has a cat in it...) >>>>>> >>>>>> No such trickery is required. Consider the experiment where the >>>>> subject >>>>> is a computer program and the clairvoyant is you, with the program's >>>>> source >>>>> code and inputs. You will always know exactly what the program will do >>>>> by >>>>> running it, including all its deliberations. If it is the sort of >>>>> program >>>>> that decides to choose both boxes it will lose the million dollars. The >>>>> question of whether it *ought to* choose both boxes or one is >>>>> meaningless if >>>>> it is a deterministic program, and the paradox arises from failing to >>>>> understand this. >>>>> >>>> Not trickery, how dare you?! An attempt to give a meaningful answer >>>> which >>>> actually makes something worthwhile from what appears to be a trivial >>>> "paradox" without any real teeth. >>>> >>>> But OK since you are determined to belittle my efforts, let's try your >>>> approach. >>>> >>>> 1 wait 10 seconds >>>> 2 print "after careful consideration, I have decided to open both boxes" >>>> 3 stop >>>> >>>> This is what ANY deterministic computer programme (with no added random >>>> inputs) would boil down to, although millions of lines of code might >>>> take a >>>> while to analyse, and the simplest way to find out the answer in >>>> practice >>>> might be to run it (but each run would give the same result, so once >>>> it's >>>> been run once we can replace it with my simpler version). >>>> >>>> I have to admit I can't see where the paradox is, or why there is any >>>> interest in discussing it. >>>> >>>> It's probably not a true paradox, but why it seems like one is that >>> depending on which version of decision theory you use, you can be led to >>> two >>> opposite conclusions. About half of people think one-boxing is best, and >>> the >>> other half think two-boxing is best, and more often then not, people from >>> either side think people on the other side are idiots. However, for >>> whatever >>> reason, everyone on this list seems to agree one-boxing is best, so you >>> are >>> missing out on the interesting discussions that can arise from seeing >>> people >>> justify their alternate decision. >>> >>> Often two-boxers will say: the predictor's already made his decision, >>> what >>> you decide now can't change the past or alter what's already been done. >>> So >>> you're just leaving money on the table by not taking both boxes. An >>> interesting twist one two-boxer told me was: what would you do if both >>> boxes >>> were transparent, and how does that additional information change what >>> the >>> best choice is? >>> >> If both boxes were transparent, that would screw up the oracle's >> ability to make the prediction, since there would be a feedback from >> the oracle's attempt at prediction to the subject. The oracle can >> predict if I'm going to pick head or tails, but the oracle *can't* >> predict if I'm going to pick heads or tails if he tells me his >> prediction then waits for me to make a decision. >> > Why not? If the oracle has a complete and accurate simulation of you then > he can predict your response to what he tells you - it's just that what he > told you may then no longer be truthful. Suppose he tells you he predicts > you'll pick tails, but he actually predicts that after hearing this you'll > pick heads. It just means he lied. It's somewhat like the unexpected > hanging problem, if you reason from the premises given and you reach a > contradiction then the contradiction was implicit in the premises and no > valid conclusion follows from them. >
If the oracle lies then he controls the inputs and can still make a prediction. The difficulty arises when he tells the truth, which is effectively what happens with transparent boxes. I think then the answer for the oracle is then indeterminate since you can always do the opposite of what he tells you; although a super-oracle outside the system might still be able to predict what will actually happen. -- Stathis Papaioannou -- 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/d/optout.
