Imagine that someone offers you a deal: a quantum random number generator will randomly display a message “WIN” or “LOSE”. (The chances of each result are non-zero.) If it is “WIN”, you will get million dollars. If it is “LOSE”, you will be immediately killed painlessly in your sleep.
According to the Quantum Immortality hypothesis, you should take this deal, because in all quantum branches where you will exist you get million dollars, and the quantum branches where you don’t exist are simply not your problem. So there is no downside to this bet. Do you agree?
If you said yes, imagine that there are two quantum random number generators available. One of them displays “WIN” with probability 99.99% and “LOSE” with probability 0.01%. The other displays “WIN” with probability 0.01% and “LOSE” with probability 99.99%. Do you have any preference at all about which of these two generators should be used in your case?
According to the Quantum Immortality hypothesis, you should take this deal, because in all quantum branches where you will exist you get million dollars, and the quantum branches where you don’t exist are simply not your problem.
Why do people quietly assume QI implies a save-point-like functionality, where you continue from a convenient-to-your-conscious-mind position?
QI doesn’t imply you only wake up in the universes where you weren’t killed. You also wake up in universes where the killing was botched. You also wake up in universes where the scenario was a lie and you’re now a slave in an underground mine. You also wake up in a myriad of quantum branches where your mind has fractured into discontinuity and exists, a fraction of a second each, across myriad distances of space and time, forming briefly as matter and energy randomly, and against great odds, configure into something that runs your mind for a brief period of time before falling back into chaos. And all of that is assuming your mind actually -stopped- running while you were asleep; the kinds of universes you wake up into if your mind -didn’t- stop running while you were asleep should be expected to be much uglier.
Your mind appears, on average, wherever your mind is most likely to appear, not where you would naively expect your mind to appear, nor where it would be convenient for your mind to appear.
Given that:
If you said yes, imagine that there are two quantum random number generators available. One of them displays “WIN” with probability 99.99% and “LOSE” with probability 0.01%. The other displays “WIN” with probability 0.01% and “LOSE” with probability 99.99%. Do you have any preference at all about which of these two generators should be used in your case?
Given that somebody says yes, they should have a preference, yes. The more likely winning is, the -less- likely they are going to end up in a hellish existence they didn’t anticipate at all.
According to the Quantum Immortality hypothesis, you should take this deal, because in all quantum branches where you will exist you get million dollars, and the quantum branches where you don’t exist are simply not your problem. So there is no downside to this bet. Do you agree?
Right now you’re on some branch B. It will have lots of future “descendants”; call them B1, B2, etc. All of them are possible futures for you-right-now-on-branch-B.
Viliam’s account of a believer in “quantum immortality” has them reasoning as follows: If I take the deal then in (let’s say) branches B1, B3, B5, etc., I will be $1M richer and in branches B2, B4, B6, etc., I won’t exist. I only care about branches in which I exist, and in all of those taking the deal leaves me $1M richer. Therefore I should take the deal.
There’s no suggestion that our hypothetical quantum-immortalitarian cares (or should care, or thinks she should care) about her analogues on “parallel” branches. Only that she cares about the branches that are possible futures for her.
Not caring about those branches surely just means not caring about your own future.
(Note: of course thinking about “branches” as nice neat discrete things to which one can give serial numbers is wrong, but it makes for simpler exposition.)
Yeah, perhaps writing about “you in other branches” was misleading. Let me try to clarify: a branch does not start with the point of departure, in every Bn there’s also the entire B before that point. So you basically exist in every branch, only that in some branch you stop existing sooner. ‘Not caring about other you’ means in this case that I see no difference between the quantum russian roulette and a purely classical one: would you play a classical russian roulette with half the bullets, for a megadollar? I certainly wouldn’t, so there’s no reason I would play the quantum one.
I also see no important difference between classical and quantum Russian roulette, but I still don’t think I understand how what you’re saying about branches relates to that.
In any case, unless I’ve misunderstood Viliam’s original comment, it seems like you and he are in agreement; his argument is meant as an anti-QI intuition pump and is really only directed at those who (unlike you, unless I’m confused) endorse QI.
On a second thought I too couldn’t find any possible difference between quantum and classical roulette, if your preference is not shifted by the measure of the branches where you’re not dead.
Given that I was 100% certain that Quantum Immortality is true and works like that, given that I was 100% certain of the other implicit assertions (e.g. you don’t go to hell for committing suicide), and given that I was 100% certain of the mechanism working as stated (so that the killing mechanism doesn’t maim rather than kill you or accidentally cause suffering in the process), then I would take the bet and have no preference over which of the two generators to use.
I am not, and never will be, 100% certain of any of those things.
My intuition is that this is one of those cases where given t “evaluation on the left side of t” and “evaluation on the right side of t” give different results. It seems to me that at any given time decision is made about future actions (and not the past), thus “evaluation on the left side of t” seems to be more important and it is the one that makes me reluctant to play this game. It seems to me that using “evaluation on the right side of t” (in cases where they differ) might give some strange results, e.g. murder having no victims.
It seems that left side of t and right side of t differs whenever there is different number of people on both sides. E.g. if you make an exact copy of a person and their entire memory, the “left identity” and “right identity” (perhaps there are better terms) intuitively seem to become two different things.
Imagine that someone offers you a deal: a quantum random number generator will randomly display a message “WIN” or “LOSE”. (The chances of each result are non-zero.) If it is “WIN”, you will get million dollars. If it is “LOSE”, you will be immediately killed painlessly in your sleep.
According to the Quantum Immortality hypothesis, you should take this deal, because in all quantum branches where you will exist you get million dollars, and the quantum branches where you don’t exist are simply not your problem. So there is no downside to this bet. Do you agree?
If you said yes, imagine that there are two quantum random number generators available. One of them displays “WIN” with probability 99.99% and “LOSE” with probability 0.01%. The other displays “WIN” with probability 0.01% and “LOSE” with probability 99.99%. Do you have any preference at all about which of these two generators should be used in your case?
Why do people quietly assume QI implies a save-point-like functionality, where you continue from a convenient-to-your-conscious-mind position?
QI doesn’t imply you only wake up in the universes where you weren’t killed. You also wake up in universes where the killing was botched. You also wake up in universes where the scenario was a lie and you’re now a slave in an underground mine. You also wake up in a myriad of quantum branches where your mind has fractured into discontinuity and exists, a fraction of a second each, across myriad distances of space and time, forming briefly as matter and energy randomly, and against great odds, configure into something that runs your mind for a brief period of time before falling back into chaos. And all of that is assuming your mind actually -stopped- running while you were asleep; the kinds of universes you wake up into if your mind -didn’t- stop running while you were asleep should be expected to be much uglier.
Your mind appears, on average, wherever your mind is most likely to appear, not where you would naively expect your mind to appear, nor where it would be convenient for your mind to appear.
Given that:
Given that somebody says yes, they should have a preference, yes. The more likely winning is, the -less- likely they are going to end up in a hellish existence they didn’t anticipate at all.
If you die, your family/friends would be sad.
Assuming that you care about all the ‘you’ in other branches. I don’t, so I wouldn’t take the bet.
At least one of us is confused.
Right now you’re on some branch B. It will have lots of future “descendants”; call them B1, B2, etc. All of them are possible futures for you-right-now-on-branch-B.
Viliam’s account of a believer in “quantum immortality” has them reasoning as follows: If I take the deal then in (let’s say) branches B1, B3, B5, etc., I will be $1M richer and in branches B2, B4, B6, etc., I won’t exist. I only care about branches in which I exist, and in all of those taking the deal leaves me $1M richer. Therefore I should take the deal.
There’s no suggestion that our hypothetical quantum-immortalitarian cares (or should care, or thinks she should care) about her analogues on “parallel” branches. Only that she cares about the branches that are possible futures for her.
Not caring about those branches surely just means not caring about your own future.
(Note: of course thinking about “branches” as nice neat discrete things to which one can give serial numbers is wrong, but it makes for simpler exposition.)
Yeah, perhaps writing about “you in other branches” was misleading. Let me try to clarify: a branch does not start with the point of departure, in every Bn there’s also the entire B before that point. So you basically exist in every branch, only that in some branch you stop existing sooner.
‘Not caring about other you’ means in this case that I see no difference between the quantum russian roulette and a purely classical one: would you play a classical russian roulette with half the bullets, for a megadollar? I certainly wouldn’t, so there’s no reason I would play the quantum one.
I also see no important difference between classical and quantum Russian roulette, but I still don’t think I understand how what you’re saying about branches relates to that.
In any case, unless I’ve misunderstood Viliam’s original comment, it seems like you and he are in agreement; his argument is meant as an anti-QI intuition pump and is really only directed at those who (unlike you, unless I’m confused) endorse QI.
On a second thought I too couldn’t find any possible difference between quantum and classical roulette, if your preference is not shifted by the measure of the branches where you’re not dead.
Given that I was 100% certain that Quantum Immortality is true and works like that, given that I was 100% certain of the other implicit assertions (e.g. you don’t go to hell for committing suicide), and given that I was 100% certain of the mechanism working as stated (so that the killing mechanism doesn’t maim rather than kill you or accidentally cause suffering in the process), then I would take the bet and have no preference over which of the two generators to use.
I am not, and never will be, 100% certain of any of those things.
My intuition is that this is one of those cases where given t “evaluation on the left side of t” and “evaluation on the right side of t” give different results. It seems to me that at any given time decision is made about future actions (and not the past), thus “evaluation on the left side of t” seems to be more important and it is the one that makes me reluctant to play this game. It seems to me that using “evaluation on the right side of t” (in cases where they differ) might give some strange results, e.g. murder having no victims.
It seems that left side of t and right side of t differs whenever there is different number of people on both sides. E.g. if you make an exact copy of a person and their entire memory, the “left identity” and “right identity” (perhaps there are better terms) intuitively seem to become two different things.