Quantum Immortality

hadeka

Hi all

Quantum Immortality: http://en.wikipedia.org/wiki/Quantum_immortality

I could not understand it ... does it prove life after death ?!!!!

DaveC426913

It is merely a thought experiment which explores the implications of the MWI of QM. It proves nothing; it is merely conjecture.

In a nutshell: If the Many-Worlds Interpretation is the correct** interpretation of QM, then it could be argued that this implies immortality of a sort. If all possibilities of a life-of-lethal-experiences create their own universe, then there must be at least one universe in which the experimenters survived every near-death experience, therefore is, in a sense, immortal somehwere out there in the multiverse.

**BTW, it can't be proven or disproven, so it is philosophy, not science.

Fredrik

It certainly doesn't prove life after death.

If you try to kill yourself, there's always some probability (very tiny, but still >0) that you will fail. In quantum mechanical terms, the wavefunction of the universe (a concept that only makes sense in the many-worlds interpretation) will evolve into a state that's a superposition of you being dead and you being alive. In the many-worlds interpretation, the wave function doesn't "collapse" into one of those two options, and since the theory doesn't suggest that one of the options is more "real" than the other, the natural conclusion would be that they're both real.

Now the idea is that from your own point of view, 100% of the times when you ask yourself "am I still alive", you are alive. And some people claim that this means that from your point of view you will always fail.

This sounds like complete BS to me. In fact it's a bit too obvious that it is. It makes me suspect that I have misunderstood something.

Edit: What I said in the first long paragraph is a valid point about the MWI. I don't think that part is BS. The MWI does imply that some version of you will always continue to exist after a suicide attempt. This version of you won't notice anything different about the world around him. He will just think that the gun jammed or whatever. The part I think is BS is that stuff about "your point of view", but maybe that wasn't part of the original argument. It could be a misinterpretation that was introduced later (possibly by Tegmark).

xantox

Which is the exact problem you have with this part?

Fredrik

Consider e.g. the statement "From Bob's point of view, there's a 100% probability that he will fail". Suppose that he goes out in the desert and kills himself by shooting himself in the gut. It will take a few days to die, but he doesn't care since according to the argument there's a 100% probability that the gun will fail.

That sounds intuitively insane, but that may not be the best counterargument. I'll try to do better. The argument isn't completely crazy, but it seems to me that the words "Bob" and "he" are references to different things. "Bob" is clearly a reference to the guy who's still alive after the suicide attempt. So Bob is a physical system that doesn't include a rotting corpse. That's why it's not the same physical system that tried to kill itself. I'll try to make that more clear.

Think of Bob's initial state as a subsystem of the universe, isolated from the rest. The "rest" includes the gun. When he tries to kill himself, he interacts with the rest of the universe. The initial state of the universe is |Bob alive>|gun loaded> and it evolves into |Bob still alive>|gun jammed>+|Bob dead>|gun discharged>.

The whole point of the MWI is the logical consequence of the fact that the final superposition represents the same physical system as the initial state: We know that one of the terms represents reality, but since there's no indication whatsoever in the theory that one of the terms is of a different kind than the other, we are forced to conclude that the theory's prediction is that there are two realities.

Recall that the word "he" refers to the physical system that tried to kill itself. According to the above, that system now consists of a guy who's still around to ask questions and a rotting corpse. So "he" doesn't refer to the same thing as "Bob".

(The reason I colored the word Bob red everywhere except in the third paragraph is that in the third, it refers to the complete physical system, not just to the survivor. All the red Bob's above are references to the survivor only. The blue Bob in the third paragraph is a reference to the corpse only).

xantox

I agree on the first two statements, but disagree on the last, ie. we should not equate "he" to some particular physical system / body before the experiment, but rather to some abstract property of it and its motion which corresponds to "being Bob". That is why the argument can work subjectively and why it can be said that "he" is Bob.

The desert example seem also not applicable, as the "dying Bob" state should not be allowed in the superposition, eg. the triggered device should rather annihilate the body, on a shorter timescale than the consciousness response time.

We had a similar discussion with vanesch on this other thread, http://www.physicsforums.com/showthr...=198794&page=2 (posts 26-31) and similarly I didn't find at the end any basis allowing for Bob to decide he is located in some particular body rather than another.

Naty1

Nobody understands exactly what either life or death is either...Just because you "die" in this universe is no proof you have "disappeared" from an infinity of other universes, if they exist. And you are "made from the stars" and your atoms will eventually be returned there to be recycled yet again, so you disperse but are in a sense immortal...

Then there is the issue of the "soul" which may or may not be essential to life...and if it is how that gets "recycled"....if it does....

vanesch

The problem with this quantum immortality is that it makes an extra hypothesis which is a priori not necessary in MWI.

In MWI, you have to distinguish between the subjective "you" and the material body support (the physical degrees of freedom of the body material). In more classical approaches, we associate a single subjective you with a single material body, so there's no point in distinguishing between both. The association is between a subjective experience, and a number of material degrees of freedom (a set of particles, say). The (single) state of that set of particles determines the subjective experience that goes with it.

In MWI, with a single set of material degrees of freedom (a single set of particles) corresponds a single quantum state (its wavefunction) but which is split into many different states (each entangled with a different state of "the environment"), and it is postulated that to each of these different quantum states (parts of the overall state) corresponds (at least one) subjective experience. So this time, to the same set of physical degrees of freedom (to the same set of particles) correspond miriads of subjective experiences. However, "you" are just one of them, and the link between subjective experience and the material world is given by the MWI equivalent of the Born rule: the probability for "you" to experience a state |psi5> is given by the relative norm of |psi5> to the other states of the same material degrees of freedom in the overall quantum state.

But now, the question is, what happens upon time evolution ? Are "you" redistributed each time ab initio over all the different quantum states that correspond to the evolving overall state, or is there some "continuity" ? That is, do you incessantly experience from moment to moment, totally different worlds, or do you "stick to your world" ? I prefer to think you do, but that's just my personal choice. That means that upon each quantum splitting of YOUR branch, you go randomly into one of the emerging branches, with probabilities given by the Born rule for THIS branching. You never visit other, parallel branches again. But that's just my favorite. The "other" branch(es) that are split off are then assigned to "new" subjective experiences, but you just continue to trace out your path through the arborescence of branchings. And now you can see it how you want: if some branches cannot support any subjective experience anymore (your body state is dead), and "you" happen to "draw" it, then I could say that this terminates "your" subjective experience. You are dead now. In the next door branch, a "new" subjective experience is created, but it is not the original "you", and so you don't experience that, in the same way as you don't experience the "other" branch in a non-fatal splitting. If that's the "rule", then quantum immortality doesn't work. You are just dead, and in a parallel world, a copy of you survives.

However, you can modify the rule, and say that you only draw from the "surviving" branches. And then, quantum immortality works the way it is displayed.

In the case you don't like to keep "subjective continuity" but redraw each time from scratch, the same can be said: or you draw only from "living states", in which case quantum immortality happens of course. Or you draw from living and dead states, in which case you have also a chance to be dead. The strange thing is that the next moment, you might draw again a branch in which you live...

So quantum theory, even in the MWI interpretation, doesn't lead unavoidably to quantum immortality.

However, if it is the case, then this could be the basis of an "improbability drive" (from Douglas Adams fame...)

xantox

The problem is that it is not possible to localize one instance of "you" when there are many identical ones. So, upon each quantum splitting of YOUR branch, "you" don't go randomly into ONE of the emerging branches, but you go into MANY emerging branches. It's only upon subsequent branching that each instance will become slightly different. But after the initial branching which triggered the device, both "you" are the same entity.

For this reason, just before a branching, the "you" before the branching cannot consider that the "left path" will be any more or less valid than the "right path" in order to become "your path".

As a consequence, after the branching, and since we put a roadblock on the left path, the "you" who took the right path must both consider to have survived the experience and to be still on "his path".

pallidin

My understanding is that this is like a "catch-22"
Only the actual participant can experience the potential quantum immortality effect as opposed to anyone else or any external instrumentation to verify its occurrence.

Thus, though a 'thought experiment", it is intrinsically un-verifiable under those terms.
This leads me to question the value of that and similar types of "thought experiments"

pallidin

Having given this more thought, I wonder if there is a way to construct a MWI quantum immortality experiment that can be verifiable.

The concept of the "quantum eraser", for example, encourages me to think that exotic experimental techniques could be developed to explore brazen ideas. In this case, though, I'm not sure. It's a tough one.

Fredrik

We seem to agree that the time evolution of the state of the universe around the time of the suicide attempt can be described as

|Bob alive>_X|gun loaded>_Y ---> |Bob still alive>_X|gun jammed>_Y + |Bob dead>_X|gun discharged>_Y

where X is a physical system that includes the original Bob, and Y is everything else, including the gun. Our disagreement seems to be that I'm saying that it's the physical system X that's trying to kill itself, while you're saying that it's a proper subset of the Hilbert space of possible states of X that tries to kill itself.

Explanation of that last part: "Being Bob" is a property of the physical interactions in his body. The interactions are described by his quantum state, which is represented by a vector in the Hilbert space H(X) of possible states of the system X. So to say that the system is "being Bob" is to say that the state vector belongs to a particular proper subset B of H(X).

I think it's pretty strange to think of Bob as a subset of states rather than as a physical system. If we do, we're introducing the problem of what other subsets might be relevant. Do we e.g. have to define a subset B' of B that consist of states that represent a guy that's "alive and not dying"? Then how do we define "not dying"?

If this is what the quantum suicide argument is about, I think they should have made that clear from the start. (Maybe they did. I've only read one article about it, and I don't remember the details).

I disagree about the need for the whole thing to be over quickly. The only reason I can think of to treat the suicide attempt as effectively instantaneous would be to hide some of the issues, in particular the need to consider proper subsets of B.

It's obviously impossible to make the interaction instantaneous. It is however possible to make its duration so short that there's no time for any real thoughts, but that doesn't change anything important unless we really believe that there's something fundamentally different about a physical system that's conscious. What about someone who's asleep, in a coma, or clinically dead during brain surgery?

I was wrong to say that quantum immortality claims that the dying Bob is right to say that there was a 100% chance that he would fail. I'm not sure what the immortality argument really says though.

Just as a reference, if this discussion continues, this is how I'd like to express the time evolution of the state of the universe when we let it take a while:

|Bob alive>_X|gun loaded>_Y ---> |Bob not dying>_X|gun jammed>_Y + |Bob dying>_X|gun discharged>_Y

---> |Bob still alive>_X|gun jammed>_Y + |Bob dead>_X|gun discharged>_Y

xantox

I would certainly not include my hair in the description of myself, especially since I tend to lose many of them. But also in my brain, there must be a majority of degrees of freedom not bearing the "being myself". While we do not have the necessary knowledge to identify them, there is certainly such a subset. "Not dying" can be defined for the needs of the thought experiment in a larger way as simply not disrupting the whole body eg by nuclear fission.

The basic reason for the quick timing is making sure that no instance of the experimenter will experience the forbidden branch, or that will ruin the whole experiment. The probability that Bob can observe that branch must be zero. This includes not hearing a bang, or not suffering for a few seconds. If the timing is quick, as you said no conscious experience can develop, then ensuring that the "myself" system can't experience his own presence in the forbidden branch. Possibly there are other problems in cutting the brain evolution at any random time, but one supposes they can be solved so that the end of the "myself" evolution in the forbidden branch can be perfectly plugged in the evolution of the other branch without any discontinuity. Indeed performing the experiment while sleeping could be an additional measure, though the experimenter could prefer to be conscious to make sure the device is operated correctly.

Fredrik

That's actually not the kind of subset I'm talking about. I'm not talking about splitting the Hilbert space into subspaces that represent different degrees of freedom, so I am including the hair or whatever in the system X. I'm talking about a subset B of vectors in H(X) that represent versions of Bob that we would consider conscious. Note that for every vector in B that e.g. has the spin of some electron in your foot pointing up, there's another one with that spin pointing down.

Now correct me if you think I'm wrong, but I think that's the sort of subset of H(X) that would represent the "being Bob" property of the system X.

But why? Why not just wait until sufficient time has passed, so that we know that the system is described by the last superposition in #12.

I really don't see why a "conscious" physical system should be treated differently than one that's not.

Fredrik

You didn't make it really clear what the extra hypothesis is. Can you express it as a sentence?

vanesch

That the "choices" of a subjective experience are limited to the "living" branches and cannot end up in a "dead" branch.

It makes the difference between: "I got a twin, but he died" and "I got a twin, but I died".

What quantum immortality actually needs, is: "I got a twin, and upon the death of my body, I became my twin".

vanesch

Observationally, you're right. But imagine that there is a perfect twin of you, somewhere in a perfect copy of your environment. That's a very similar situation, and your argument would also apply: there's no distinguishing between that twin of yours, and yourself. Hey, maybe there IS such a place right now. Imagine that the next split second, that twin is killed. You're not. You're not even aware of that. Now, imagine again that the next split second, YOU are killed. Do you really think that "you" still live on as your twin ?

Yes, but the hypothesis here is that your subjective experience can jump from the left to the right. That's the extra hypothesis.