Quantum Immortality Is Useless
And now for something completely different.
Quantum immortality is a controversial speculation supported by a handful of quantum physicists and abused by many science fiction authors. In essence, it states that the many-worlds interpretation of quantum mechanics implies that conscious beings are immortal. In this post I’ll attempt to show that quantum immortality is completely useless, even if it really works as described.
You Are The Cat
The idea of quantum immortality was derived from the quantum suicide thought experiment, which is a close cousin of the famous “Schroedinger’s undead cat” experiment. Instead of a cat in a box + poison triggered by radioactive decay, the quantum suicide thought experiment involves a mad scientist and a lethal weapon (which I will refer to by the unscientific name “gun” from now on). This gun has been cleverly modified so that when one pulls the trigger, it will measure the spin direction of a passing photon and fire only if the direction is, say, “up”. This amounts to a 50% chance of the gun actually firing.
So the mad scientist points the gun at his head and pulls the trigger. Depending on the spin direction of a random photon, either the gun fires and kills the scientist, or it just goes “click” and leaves him unharmed and free to go on doing what he must because he can.
According to the Everett many-worlds interpretation, each run of this experiment will cause the Universe to branch into two alternate world-lines – one where the scientist lives, and one where he dies. However, the scientist will only experience the world where he survives, because he obviously won’t have any conscious experiences in the worlds where he’s dead. Even if the experiment is repeated an arbitrary number of times, the scientist will only experience the world(s) where he survives.
From the point of view of the surviving copy (or copies), he’ll be immortal.
That’s the essence of the quantum immortality argument. You can’t experience the world-lines where you die, so you can expect to experience only the one where you survive – indefinitely.
Who Wants To Live Forever?
As a matter of fact, I do. Unfortunately, quantum immortality is unlikely to help me achieve eternal life. In fact, the concept is pretty much useless, except as a quaint “what if?” brain-tickler. Lets look at that quantum suicide thought experiment again :
Gun stock photo credit: Fastfood @ sxc.hu
If the many-worlds interpretation is correct, there will indeed always be at least one world-branch that avoids the fatal outcome, one copy of you that survives. Even if you get thrown into an active volcano, there will be an Everett branch where Harry Potter swoops in at the last moment and carries you to safety on his magical flying broomstick (depending on your cultural preferences, this may or may not be worse than being burnt alive by magma). The proponents of quantum immortality will point at that lucky branch and clap their hands giddily.
But – and here’s the kicker – you don’t get to choose in which Everett branch you end up. Yes, there may well be a world-line or two where you live forever, but what are the chances that you – this particular version of you, reading this blog post – are in that world-line? It would be extremely generous to say they’re rather slim.
As for the “you can only experience the worlds where you survive” thing, it’s just plain fallacious. It’s like a sales-person asking you “Will you be paying by check, cash, or credit card?” when you haven’t even decided if you actually want to buy their product. The question intentionally leaves out a valid option – not buying anything – and tries to make you pick one of the options most beneficial to the sales-person.
Similarly, the idea that you can only experience the Everett branches where you are alive incorrectly implies that being dead is somehow not a valid option. Unfortunately, it is. Magic aside, there is no overriding reason why a conscious creature couldn’t just… die, and thus stop being conscious (though that might change sometime this century).
Even if Everett was right and there are numerous alternate worlds, and in some of them a version of you is subjectively quantum-immortal, that doesn’t change your chances of survival one bit.
Related posts :
Subscribe via RSS
Get Updates by Email
[…] Quantum Immortality Is Useless […]
http://www.msnbc.msn.com/id/12425803/
Any conscious version of myself in the future will remember being me now. The me now becomes all future possible versions of me. If only one of those future versions survives, it will still be ‘me’ and remember a seamless existence from being me now, until me then.
What you’re saying is that the many versions of myself are already separate entities in their own closed systems, evolving along their own paths. This isn’t the case. If there is 50% chance of me dying in my sleep tonight, that doesn’t mean theres a 50% chance of me waking up in the morning depending on what universe I’m in now. It means there’s a 100% chance of me waking up, as I’m only going to be conscious in the universe where I didn’t die. I’m not yet in either of those universes because the quantum states which result in the 50/50 split have not yet decohered, and until they do, the dead me and the alive me in the morning are both the current me now. You’re implying that in my circumstance, there are 2 versions of myself writing this comment, one which will be alive tomorrow and one which won’t.
I may make the decision to switch on my light in the morning, electrocuting myself to death. I may decide to get out of bed without touching the light, and I will live another day. Until I make that decision, those 2 seperate realities have not split, and I am in neither until the moment occurs. I will therefore only find myself in the reality where I survive.
Upon further reflection, I think it boils down to the question whether you accept the implied premise of quantum immortality that any conscious entity sufficiently similar to your current self counts as “you”.
If you do, then quantum immortality certainly works. If you don’t, there will still always exist at least one copy of yourself – but it won’t necessarily be “you”.
Any future conscious entity which remembers being me will BE me. That’s all we have to define ourselves, our memory of ourselves. All future multiple-worlds versions of myself will be the same me now, and remember being me now. Every version of myself is equally me, as there is nothing to seperately define the ‘one’ me from the other versions.
Fantastic topic, but normally I don’t agree with it.
The point is that an exact copy of you IS you. Imagine a teleportation device that destroys your body at point A and reassembles an exact replica from a store of atoms at point B. Now imagine that you’ve used this device many times to transport yourself. Some metaphysician somewhere might tell you that you’re not really you, but as a practical matter, everytime you’ve used the device it’s worked. You step in at point A and step out at point B and you’re still you.
If MWI is correct, the only way quantum immortality can be incorrect is if consciousness depends on a soul beyond the material structure of the brain.
On the other hand…..
Say you can set the teleporter to send you to two locations — points B and C. In the past, when you set the destination to B, you came at at point B. When you set the destination to C, you came out at C. What happens when you set the destination to B and C?
From an objective perspective there’s no paradox at all. One copy of you arrives at point B and reports “I set the destination to B and C, and I arrived at B.” The other copy reports “I set the destination to B and C, and I arrived at C.”
But from a subjective point of view it’s very strange. You’re in the booth, you set the controls to “B and C”, and you’re about to click “Send”. You’re bound to wonder — where am “I” going to end up? Best you can come up with is it’s a coin toss… but it’s not.
To increase the stakes a bit, imagine there’s a madman at point C waiting to chop you into pieces with a machette. The controls are set to “B and C”. Now click “Send”.
English isn’t really suited for discussing multiple instances of oneself.
An exact copy of me is A me, but it is not necessarily THE me. Destructive teleportation kills the previous me and creates a new me. Even if the two copies are atom-for-atom equivalent, they’re not the same because they’re causally disconnected.
I would never take use a destructive teleport willingly. Wll, there can be some extraordinary circumstances where I would, but that’s beside the point.
But what about the fact that our cells are always dying off and getting replaced? The atoms that make up our bodies now are not the same atoms that did 10 years ago, so why should one worry about replacing them all at once?
The key point here is that natural cell replacement is gradual. The chain of existence, the dynamic phenomenon that one calls “me” is unbroken and continuous. For example, if some superintelligent and benevolent AI offered to gradually replace all of my cells with more resilient artificial ones, I would agree.
Look at it as a tree structure. The root node is “You Now”, which branches into child nodes, and the child nodes themselves branch out, and so on. The many-worlds interpretation says that none of the paths through the future branches is preferred. There’s no single path that corresponds to the “real you” — they’re all equally actual. There’s no reason for any particular universe to be “more real” than any other.
The reason you have the strong impression that your consciousness does correspond to a particular path is that the “You Now” node does correspond to a particular path THROUGH THE ANCESTOR NODES. But “You of the Future” is a tree, and every path actually happens. There’s no rolling of the dice to determine which path gets actualized, although there’s a strong impression that there is, because “You Now” does correspond to a particular path through the past. In a strange way, it’s as if “You Now” is the given, and it always persists.
I get that. It’s the reason I distrust quantum immortality – you can’t claim that only the branches where you remain alive and well are “real”. Some of the future you will survive, but some will also die.
Now in your teleportation example, the situation seems different. The “You Now” node is cut off and discarded, and somewhere else a new tree of possibilities is created – each branch similar to, but no longer connected to the original “You”.
Uhmmm… “You Now” would be you looking down at the teleporter’s Send button. B and C are the 2 child nodes. In both cases, “You Now” is continually vanishing into the past, or rather, continually transforming into memories. The difference is that the You’s at B and C could later meet each other.
er… unless of course You at C gets chopped into pieces with a machette.
Unless you can somehow go back in time, the fact that the momentary snapshot titled “You Now” still exists somewhere in the past is not all that valuable. It’s still cut off.
It appears you are connecting the nodes using memories. I connect them using both memories and low-level physical states (e.g. atom positions relative to the planet’s surface), so in my view B and C don’t count as full continuations of “You Now”.
Yeah, I see your point — it comes down to the mind/body problem. But in the case of MWI-style splitting, where there’s a new you in universe B and a new you in universe C, which one gets the real, original atoms?
Okay, I just answered my own question! They both still have their original atoms because the splitting process doesn’t create new matter. B and C are universes with initial conditions so nearly identical that they were functionally identical for 15 billion years before they finally diverged. (“Functionally identical” means a 1-to-1 pairing exists for all particles, despite the fact that two particles in a given pair might be in slightly different positions.)
…so, come to think of it, the problem of “multiple You’s” is there even before the split. They’re all functionally identical, but which one is ME?
I’d say “the” You is the one that is currently thinking that question. In a MWI setting it may be impossible to accurately identify a particular copy since you’re branching all the time and you don’t have access to an outside reference frame from which to observe all the branches. But you really don’t need to – I think it’s sufficient to know that you’re one of the (possibly very many) instances that share a particular past.
I once read an essay on decision-theory that worked around this problem by saying you should act as if you were controlling all of your MWI-instances at once. Even though you only physically control one, making decisions as if you were all of them is supposed to work out best “on average”.
In an MWI context, the 2 ‘yous’ are in fact composed of the same atoms. The ‘yous’ at B and C correspond to decohered copies of the ‘you’ now, which currently exists in a superposition of states. This is the foundation of MWI.
The teleportation thought experiment is a good one, and serves to put decoherence into a context which most people will understand to a greater degree than they would initially. What’s important to remember is that the ‘you’ at the starting point, A, is in a superposition of states. These superpositions of states evolve deterministically along different worldlines. Neither worldline cointains a more real ‘you’ than the ‘you’ at the starting point. There is no original version of ‘you’ to begin with. ‘You’ are a wavefunction the same as a photon, though obviously much more complex.
Again, the example of teleporting yourself to 2 different places and assuming theres a 50/50 chance of ending up at B is an excellent analogy. You end up deterministically at both B and C, but the ‘yous’ at both locations think they’ve ended up at their successive destinations quite randomly. The limitation of human consciousness is why we perceive the universe as being random when in fact it’s completely deterministic.
http://www.elea.org/Miracles/
If planet C promptly blows up upon your arrival, is that any different from the teleporter on planet C malfunctioning and not creating a copy of you? And is the teleporter malfunctioning any different from you teleporting yourself solely to planet B? Say the teleporter at A broadcasts a single signal which can be received by B and C. Whether you don’t instruct C to reconstruct a copy of you; or you do and the reconstruction process fails; or you do, the reconstruction process succeeds, and you immediately die, the result is the same — you transport a copy of yourself from A to B, which you’ve done many times in the past, and it’s always worked flawlessly.