Rainer Plaga: On your quantum immortality: you're right about traffic accidents and the like. But I would guess that the fact that nobody ever gets older than, say, 300 years is a deterministic law of nature, or in other words that there is no single branch in which a 300 year old James exists. Isn't that plausible?

James Higgo: The fact that nobody over 300 exists in this universe is unsurprising. But this is one of quite a few universes. And perhaps they're hiding. Will you concede there is a one in a thousand billion chance that you will live over 300 years?

Rainer Plaga: I don't know what the chance for a branch with a 300 year old James is. But how do you know it is not equal 0?

James Higgo: Anything that's possible exists, according to David Deutsch. Sorry to appeal to his authority.

Rainer Plaga: Yeah, anything *that's possible* but I guess nobody knows whether a 300 year old James *is possible* according to the laws of nature.

James Higgo: But it's possible that all the molecules in my body just rearranged themselves by chance into the form of a teapot, so it's certainly possible that you don't die.

Rainer Plaga Hello James,

I see what you are saying now. Probably you are right. Unfortunately even when we're 300 years old we will have a hard time to convince our fellow ``branchians'' that the reason for our extreme longevity is MWI! (Analogous to Max's experiment). Could there be something extremely improbable that led to the survival of the human race? (Your idea's analogy to my A-bomb experiment) An example would be: if we could prove that it is extremely improbable that a time span of 70 million years goes by without a nearby supernova that would wipe out the human race. This would be an argument in favour of MWI. This was only an example, one can show that it is quite probable that no nearby SN happens within 70 million years.

But perhaps we can find an example that works. Actually: this would also explain the absence of extraterrestrial intelligence (Fermi's paradox).

All the best Rainer

James Higgo: Some fine ideas. So I take it you're a convert, O Immortal one!

Rainer Plaga: I believe

Actually there is a corollary: It is impossible to commit suicide. Each time you try you will find that some stupid accident or strange improbable force kept you from doing it.

Rainer

James Higgo: Absolutely. And you will miraculously survive terminal cancer.

Max Tegmark Thanks for the [QTI] article. Quite interesting. In the motorcycle case, you still get a few seconds between the quantum decision and your loss of consciousness, so the surviving version of you would have to live with the distressing knowledge that the other version actually had time to become aware of its imminent death and that 3 seconds of original life actually got killed. I made the loss of consciousness virtually immediate in the gun case to avoid such unpleasantness. But you could of course easily cook up an accident scenario reducing the time delay. The ideal accident should in my opinion have * no lead time * certainty of death (you don't want to wake up as a cripple).

Max Tegmark This immortality issue has bothered me too, and a number of other people also brought it up after this article came out. I agree with you that if the argument were flawless, I should expect to be the oldest guy on the planet (at least), severely discrediting the Everett hypothesis. However, I think I've found a flaw. After all, dying isn't a binary thing where you're either dead or alive - rather, there's a whole continuum of states of progressively decreasing self-awareness. What makes the quantum suicide work is that you force an abrupt transition. I suspect that when I get old, my brain cells will gradually give out so that I keep feeling self-aware, but less and less so, the final "death" being quite anti-climactic, sort of like when an amoeba croaks. Do you buy this?

James Higgo Max, I agree this is a flaw - did your note cross with my mail asking about Alzheimer's? But I think there muse be a solution if only we understood the concept of consciousness.

James Higgo Max Tegmark, in a recent e-mail, pointed out that progressive loss of consciousness was a problem for the quantum theory of immortality. I just realised, while reading Huw Price's book, that this is an arrow-of-time issue. If you can degenerate into an amoeba, the amoeba can regenerate into you, and there will always be 'subsequent' universes in which you have recovered from whatever damage could be inflicted, to regain your consciousness. I need to read Dennett on this, but consciousness seems to be just a collection of memories, and this can always be preserved and reproduced. Any comment on these thoughts, Max?

Max Tegmark That's an interesting point. I agree that you'll make a spectacular recovery from your brain-damaged amoeba-like state in a number of branches, but just like a broken egg only comes together again in a negligible fraction of all branches, the same will apply here. So a TYPICAL expectation for my subjective future is probably that I will gradually fade away without ever making much of a comeback. Cheers, Max ;-)

James Higgo But Max, this is a very important point. If you *know* that no matter how far you degenerate there is *always* a subsequent universe in which you are 'you' again, then the immortality problem is still there. 'Typically' we disappear in a puff of smoke. So what? Weak anthropic principle - we always end up in a universe in which we exist (which normally means, one with a history which fits the laws of physics).

Gilles Henri Maybe the question of identity is not so trivial. Who is "you"? In the "everything" hypothesis (that I share with Deutsch if I understand correctly James' previous mail) only some states are connected together from "past" to "future" following a classical "consistent history" (following Griffith), defining a possible world story. The notion of "you" can be followed by temporal continuity. But it is not so obvious if you consider very rare, quantum events that branch you on completely different worlds. What allows to say "you are 'you'" during such process? You are you because of your memory of your past, which implies an arrow of time, which implies classically allowed evolution... By the way I defended on another forum (initiated by Jacques) the idea that in the "everything" hypothesis time is really not existing, since the total wavefunction of the Universe is most probably stationary. Each "instant" we observe is only a classical subcomponent of this wavefunction, apparently connected to "past" and "future" states (much like Deutsch again if I understood it correctly). So the question of one's "future" is really a question of how many "classical" states can exist and which of them do contain a "you-like" (to be defined..) with a certain age?

James Higgo Your world-view is very similar to mine. We only see an arrow of time as we are 'creatures in time'. Time is not an objective feature of the block universe. If you adopt Huw Price's Archimedean standpoint in 'nowhen' and look through Deutsch's 'multiverse' at a certain angle, you might happen (by a one in a ten to the billion chance) to string certain 'snapshots' together so you saw a 'world' evolving in time. The difference is that I am initerested in the practical question: whatever consciousness is (I'm with Dennett), can I expect to die or should I expect to be conscious for ever? I have, as yet, heard of no reason why not.

Jacques Mallah Max, you may be just one step away from seeing the error of your ways. You seem to realize that measure is the issue. Why can't you realize that even if the amoeba went one step further and disappeared completely, or if you were to consider the amoeba to be some individual other than yourself, it would not help you in the slightest to be effectively immortal? Most of your measure is in your pre-suicide perceptions.

Hal Finney It seems that there are two ways to increase the measure of copies and near-copies of yourself which have favorable experiences. One is to try to make good decisions. Some people say that in a many-worlds model decisions are irrelevant, since everything will happen. However this ignores the fact that different branches have different measures. Even though we know that all things will happen, by making good decisions we increase the fraction which have good outcomes, thereby increasing the measure of those instances of ourselves which are having favorable experiences. The other method is by killing yourself when things go wrong. This approach is more controversial, but many people seem to have an instinctive understanding of its value. People often do kill themselves when things get sufficiently bad. By doing so they are increasing the fraction of their near-copies which have good experiences. They are not thinking in those terms, but that is the effect of their actions.

Bruno Marchal I quite agree with that. But it is necessary to be cautious. Suppose that my goal is to prove Goldbach conjecture (or any unproved big mathematical conjecture). I am not very gifted in mathematics, so I decide to proceed in the following way. I use a big array of quantum particles, let us say 2^32 particles. Each one is prepared in a superposition like 1/sqr(2)(O + 1). Then I read (measure) each particle following the order given by the array, and I decode the result in the computer-keyboard-base. In case I understand what I read as a proof of Goldbach conjecture : I am done. If not, then I kill myself with a gun (let us say). Let us suppose now that there is no such proof. With MWI, I will survive ANYWAY. For example I will survive because in some "worlds" the bullet will go through by brain without affecting it thanks to the tunnel effect. Unfortunately, in the majority of the world where I survive by tunnel effect, my brain will be damaged because the probability of a clean tunnel effect is little compare to a less clean tunnel effect. In case there is a proof of G. conjecture less than 2^32 bits, then I can expect to survive in good shape in a world with the proof ONLY IF the probability to survive (even in bad shape) by tunnel effect is much less than 1/2^32. To sum up : to use succesfully the quantum suicide, you need to compare carefully the probability of your gain with the probability to survive annihilation. (and of course you must realize that according to your friends your quantum suicide is just suicide, and there are infinitely other ethical problem).

(Note : More on this in my 1998 PhD Thesis, or in my 1988 or 1991 paper, which I intend to put on a WEB page as soon as possible. In my thesis I show that the MWI is directly deductible from the thesis "I am a machine", so that empirical quantum mechanics is a confirmation of mechanism. I show that most of the qualitative aspect of quantum mechanics (indeterminism, mon-locality, etc.) are derivable fromarithmetic + mechanism. Consciousness including the appearance of matter emerge from arithmetical truth. This give a kind of "many (computational) histories, no universes" interpretation of number theory.

James Higgo Fantastic stuff! Someone who has 100% the same understanding of the situation as I do! But surely, then, Bruno, you also wonder about the issue of whether we will ever experience death. I reiterate what I discussed with Max above: no matter how 'dirty the tunnel', there will always be subsequent branches in which your brain regenerates fully. Can you shed any light on this issue? The point you make about your relatives grieving in most universes is accurate and this is the only reason I can think of (bar my genetic programming, and my genes can take a jump) for not indulging in quantum suicide.

Bruno Marchal Well ... I do think that nobody experience (absolute, first person) death. Because either we survive clinical (3d person) death, in which case we do not experience death, or we don't survive clinical death, in which case we do not experience death because to make an experience you must survive it ! (Disappointing answer I guess, it doesn't depend on MWI or Mechanism). Topologically, time-life is open, we cannot experience the border of life.

JH> I reiterate what I discussed with Max above: no matter how 'dirty the tunnel', there will always be subsequent branches in which your brain regenerates fully. Can you shed any light on this issue?

OK James, I can try. As Hal Finney put it, it is a question of (conditional) measure. Suppose you jump out of the window which is just 5 meter high. In that case, the probability you will survive is rather high. The probability you will be wounded will be rather high too. Nevertheless, the probability you will recover is not very high. I mean it is just a classical probability. The problem with the "dirty tunnel" is the fact that you survive in bad shape. And once you have survived, probabilities are classical. And, although you are right in saying that there will be subsequent branching in which you will recover fully, you cannot take that into account because your expectation for that personnal recovering is very low. ... Unless you reiterate the quantum suicide. But here, the problem is that you are taking the risk to find yourself in a so bad shape that you will not be able to reiterate the suicide. In that case, you will be stuck (for example paralysed) in a universe, perhaps for a very long time. For the long run you are surely (with Mechanism or with MWI) making a point.

JH>Another point - committing suicide is always going to be against your genetic program as your genes act as if they wanted their structure to appear in as many universes as possible. Form their point of view it is better to have you arbitrarily unhappy but existing than to have you decompose.

I think it is plausible that in most future the genes will lose the battle against other form of memes. We can hope that earth will at least be used as a kind of carbon-life and genes museum. But most practical computationalist will explore the multiverse by exchanging their expensive carbon based body for radio-waves and silicon hardware. From OUR point of view it is better to be happy without genes than unhappy with genes ...

>JH …mechanism is one valid way of looking at things but not the only…

I am indeed convinced that mechanism is not the only way to look at things. The beauty of mechanism is that with mechanism, it is even necessarily so. I mean mechanism entails the consistency of non-mechanism. A little like the fact that consistency of arithmetic entails the consistency of the non consistency of arithmetic ( a form of second Godel's theorem).

I have put the thesis on my web page. You can load it at http://iridia.ulb.ac.be/marchal