On 31 Mar 2012, at 18:24, John Clark wrote:

On Thu, Mar 29, 2012  Bruno Marchal <marc...@ulb.ac.be> wrote:

> You should care to be able to answer the simple question: "what do you expect to feel in the multiplication-movie experience"

I would expect to feel exactly the same as if duplicating chambers and multiple copies of myself were not involved.

Exactly. So you will feel to see one specific movie. That's the point. Then combinatorial analysis suggest that betting on white noise is the best choice, equivalent to betting a beam splitter splits in half sheaves of photons prepared in the relative oblique state.

If you perform whatever calculations you think are needed to produce your patented "1-view indeterminacy" I would not be one bit wiser than if I calculated probabilities using regular old conventional probability methods,

That's the point. But it works in the deterministic duplication setting *because* the probabilities appears here to be "objective" probability on different subjective experiences. The 1-view indeterminacy is only the indeterminacy, even the "Pascalian one", bearing on subjective experience. The term subjective probability would be nice if it was not already used in a context of non objective probability, like with the Bayesian approach to probability.

you've added tons of bells and whistles but for all the complex convolutions you have not added one single bit of additional information about what is likely to happen.

On the contrary, comp entails that you should expect white noice for the multiplication movie experience, that you should expect the WM duplication as equivalent with the throw of a random coin, etc.

But you don't need to agree with that analysis. You need only to agree that there is an indeterminacy on the {W, M}, or the movie, in those experiences. To reduce physics to arithmetic, will consist in reducing physics to a probability calculus bearing on a more complex domain, and we will have to leave intuition for the study of self-reference.

So I think we could move on step 4. You cannot stop in a middle on a proof on the pretext that you have learn nothing. Proofs are constructed in such a way that each step should seem easy and tautological.

> You said you favor Everett's QM.

Yes, but the universe will be the way is and doesn't care if I favor it or not, so it may or may not be true. And if Everett is right then there is a 100% probability that anything that can happen will happen, but using statistics in this way would not produce anything we could use, just as the "1-view probability" of something can not tell me anything I didn't already know.

There is a difference between anything can happen in the "reality", and anything can happen in the next moment. If Everett is right the probability must be derived from the statistics of measurements *as described by the wave evolution*. It is another debate to see if that work, as Everett claimed, and how far it relies on the 1-comp indeterminacy. At this stage of the UDA it would distract us, and we don't need the answer to proceed.

> You did not answer Quentin when he commented that with Everett the Universe is a "duplicating chamber", so that your charge again the coimp-1-indeterminacy applies to Everett QM too.

In the first place I like Many Worlds for esthetic reasons but I'm far from certain it's true.

We can never know public truth.
nevertheless I tend to believe that QM without collapse = many-worlds. That's provable if you define a world by a set of events closed for interaction. A (QM) world is a "sphere" of contagion of superpositions.

More important, in the thought experiment about the cities we have access to all the branches and can see what all the copies are doing, but that is not the case with Everett. If Everett is correct then probability is not inherent in the event itself but is just a measure of our ignorance. As you point out Many Worlds is a deterministic theory so if we had a bird's eye view of everything probability would be a useless tool because everything would have a probability of 100% or 0%.

OK, but for the probabilities, it remains the same. In the multiplication movie experience, you don't have really any practical access the other copies. The distinction you are introducing concerns what the outsider can write in his diary (the 3-view), it does not bear on the individual diary contents (the 1-views).

But I can hear you scream "but you still wouldn't know for certain if "you" will see Washington".

Why screaming? No. I am cool and ask you this in a friendly, yet firm, cool tone :)
But that was the question indeed.

I say the probability is 100% you say it is 50%, how can we tell who is correct? I say that after the experiment if I find "you" and you says that he is in Washington and only Washington then my prediction was proven correct, and I can do exactly that. If after the experiment if you find "you" and he says that he is 50% in Washington then your prediction was proven correct, but you can't find anybody like that.

You concede that you will indeed say you are in Washington and only Washington, but you correctly point out that you will also say you are in Moscow and only Moscow.

No. After the experience, both "concedes" that they feel in one city, as they describe in their respective diary, and both recognize that they don't feel to be in both city, and that they don't feel any possible sensation of their respective doppelganger. Here the 3-view, written by the observer (not going through the duplication, but observing the duplication of "you"), contains: "I see John Clark is in both cities". But the 1-views are become multiple, and one contains "I feel to be in W", and the other contains "I feel to be in W". So, it is 50%. because if you did say, in Helsinki, 100% for Washington, then, after the duplication, the guy in Moscow will understand that he was false. Like in the movie-multiplication experience, if you say 100% for the movie "flying circus", the 2^<many> -1 doppelgangers will say "I was wrong".

I agree and say that means the probability you will see Moscow and only Moscow is also 100%,

Because here you talk on the 3-view, or on the 3-view on the 1-view. You avoid again talking about the 1-view directly, which is what we need to pursue the reasoning.

but you disagree and say it is 50%.

Because we predict the experience, from the experiencer point of view. This would not make sense only if comp was false, and you tell me that you die in the process. With comp, you don't die, and you remain one and entire, so you will experience only once city.

I say the fact that you say you are in Washington and only Washington in no way weakens the claim that you are in Moscow and only Moscow.

But I feel to be in only one city.

I predict there is a 100% chance there is a 100% chance you will feel like you are in one city and one city only, and it's not a problem for me to have two yous with 100% because YOU HAVE BEEN DUPLICATED.

But n Helsinki you cannot know for sure in which city you will wake up. If you say in Helsinki that there is 100% chance that you will awake in W, and 100 chance that you will awake in Moscow, then the guy who awake in Washington must admit that he was wrong, or that he did not understood the question. because, by the definition of the 1-view (or subjectivity, consciousness, whatever) he feels only in one city. he know by quasi-direct impression that he is in Washington (resp. Moscow) and he knows only intellectually that he is also in Moscow (resp. Washington). So he might say then "Ah, that was what the question was about", and predict 50% if the experience is proposed a new time.

But exactly what does your 50% really mean? You're treating it not as a measure of ignorance

Well, you can treat it as a measure on ignorance, by using the PUP principle. Before going outside the reconstitution box, you can consider that you are already localized, and just not knowing where you are. But this need not to be precised to go to the next step.

but as if it's part of the thing itself even though it's a deterministic process, as if there is a 100% chance you will feel like YOU are 50% in Moscow and a 100% chance YOU will feel like you are 50% in Washington. And that does not correspond to the experimental results.

I have emphasized the "you" in the quote above, because those "you" are 3-view on 1-view. You are using the "bird view" on yourself. But you have to placed yourself in one of those views, or the two of them, but not simultaneously, to get the point, that for them, 100% would have been non correct. 100% in Washington is refuted by the one in Moscow, and 100% in Moscow is refuted by the one in Washington.

By definition, the 1-indeterminism bears on the 1-views.

>> My complaint is that the diaries add nothing, it's obvious that if the diaries the people remember writing are identical then the people are too, and if they aren't then the people aren't either.

> But the you contradict your statement

Did I? I hate it when that happens.

> that both the guy in W and in M have the right to say that they are the guy who was at Helsinki, which makes indeed sense with comp.

True it does make sense, the Washington guy is the Helsinki guy and the Moscow guy is the Helsinki guy and the Moscow guy is not the Washington guy and the Washington guy is not the Moscow guy. It's all very clear, but where did I contradict myself?

By saying that they are different, so that the question of probability asled in Helsinki would not make sense. But they make sense, because they are different.

>> And to add to the confusion sometimes you admit that they would feel the same, but then in your next breath you start talking about how it's identical in the "3-view" but not the "1-view".

> Could you quote me and be more precise.

Bruno, "3-view" and "1-view" are your terms, you invented them not me,

This terming is perhaps mine, and the emphasis on the importance of it too, but those notion are as older as humanity, and is reflected in natural language. Basically the 1-view is the subjective(*) experience, and the 3-view is what an outsider can "objectively" described.

(*) or conscious, but this is not needed in UDA, in fact UDA can be done with only the strong AI: then the conclusion is that machines will correctly believe that physics is a branch of arithmetic/ machine's theology.

if you did not think things could be the same in one view but not the other, if you thought the two things were identical then why the hell did you go to all the trouble to give them different names?

See above. Some people get a better understanding of the 1/3 difference thanks to the duplication experience. But in UDA the difference appears in the step two. You have to consider two classical teleportation experiences, from Earth to Mars. So you are read on Earth, annihilated on Earth and reconstituted on Mars from the info sent by some laser beam. In the first experience there are no special delays of reconstitution, in the second experience a delay of one year is introduced, and you are reconstituted one year later. I use mars, to ensure that you have no environmental clue of the time passed in your neighborhood. Do you agree that you would not been able to decide which protocol (with and without delay) has been chosen. This second experience gives a simple case where the 1-view and 3-view are different.

The 3-view are respectively, for the two experience, of the type "he is read and annihilated in Brussels, and he reappears on Mars after a short time (say)" and "he is read and annihilated in Brussels, and he reappears on Mars after a time of one year".

But the 1-view remains completely identical "I was on earth and now I am on Mars".

OK?  (I thought you already get the preceding steps).

> In the step three experience we are talking about, they will give quite different answer.

Then obviously they become different people and the thought experiment becomes rather dull.

They became different person, but they kept the memories of what they were betting in Helsinki, as you agree they are still, as person, genuine future of you. So if in Helsinki the bet was 100% washington, the guy in Moscow will understand that he aws wrong (or that he did not understood the question which was bearing on the experience as it is living now).

>> The third party does not know which one is you and you don't know either.

> The W guy know that he is a the guy in Helsinki, now instantiated in W. The M guy know that he is a the guy in Helsinki, now instantiated in M.

Yes, assuming they received information (sights sounds smells etc) from their respective cities, which in this context is all "instantiated" means.

> They both know that they are not the guy in the other city.


> They can both acknowledge the first person indeterminacy: they could not know in advance the city in which they feel now having been reconstituted.

I know with 100% certainty that I will feel to be in one city and one city only and I know with 100% certainty that the Moscow man will see Moscow because seeing Moscow is what makes the Moscow man be the Moscow man, and I know there is a 0% chance the Moscow Man will see Washington because then he'd be the Washington man not the Moscow man.

Sure. But the probabilities are asked to the Helsinki man.
If I throw a dice, I know that if I get six, I will say that there is 100% that I have six. But we ask the question before the dice is thrown, and likewize, we ask the probability on the *experience* of being in W, or M, before the duplication.

> he would not be able to predict that he will end up in M, (resp in W), although he can predict that he will end up with certainty in M or W. This shows also that when you duplicate yourself you get one bit of information. That bit of information is part of both first person experience 1p. On the contrary the 3p view does not create, or receive, one bit of information.

He will predict with absolute certainty that after the experiment is completed he will receive one bit of information indicating he is in Moscow and he will receive one bit of information indicating he is in Washington. If he can predict that now before anything was duplicated then no new information was produced or destroyed in the process and the procedure was completely reversible, and that is just what you'd expect to happen in a deterministic process like this.


> I will ask you to do the "hairsplitting" about that "YOU", that you are using here,

OK, the duplication of "you" means that this instant in time is special because this is a potential branching point, from now on if the new copy and the original body of Bruno Marchal receive different environmental input they will evolve differently, perhaps very differently, so that they become different people; but one can not say that one is more you than the other. So it is NOT a contradiction to say they are not each other but they are both you. Why is it not a contradiction? Because YOU HAVE BEEN DUPLICATED.

I never saw any contradiction there. What you say is an argument for using 50%. Given that both agree that they could not have guess in Helsinki what they are living right now.

> I will ask to give us an algorithm predicting the result of the future self-localization experience.

You agreed with me that to speak of the position of a consciousness is not productive, so all "self-localization experience" means is receiving information from Washington and Moscow, so the probability of you becoming the Moscow man are exactly the same as the probability you will receive information from Moscow. You're adding needless complications that add nothing.

The probability that I become the Moscow man is the same that you will receive the information from Moscow. That's right, but you fail to give me the probability.

> Please, don't answer me again "W and M", because we already know that a machine cannot perceive a distant environment, and that she will describe, as a result of self-localization,

I don't know what that means.


It means you look around you, may be you ask some people on the street in which city you are, and with some luck you get the answer. You might use a modern GPS device, looking at map and picture and do comparison. I use all terms in the most common sense possible in such experience. Thanks to comp, the copies survive with the knowledge of english, and with their ability to feel and they can ask people on the street, if they are lost. I give you a hint: if the technicians working in the reconstitution center talk russian, your probable self- localization is Moscow.

And in both Moscow and Washington, by comp, you can't have telepathic ability making you experience that the other doppelganger has been reconstituted or not. You will need third person clues.

> after the duplication and when opening the box, a precise city.

I predict she will see 100% of one city, you predict she will see 50% of two cities. I bet I win.

If you predict in Helsinki that you will see 100% of Moscow, do you agree that the version of you in Washington will understand that he was wrong in Helsinki? If not, it means that you are just continuing to confuse the 1-view and the 3-view, that is the subjective experience and the bodies. Or the 3-view on the 1-views (like saying: "after the duplication, you can join me in W *and* in M") and the 1-views themselves, like saying after the movie-multiplication, "yes I have seen only one movie, and the hell of predicting any bit of it, it seems to me rather random", or saying "yes, I have flying circus with the chinese subtitle of "Gone with the wind".

OK. Tell me if you get the point of step 2 and 3. And 4. Which just mixed the step 2 and 3. If you agree with David's comment, then you should be able answer step 4, if you get the step 2 point.

The step 4 is a WM duplication where a delay is introduced on only one branch, Moscow, to fix the thing. So you are duplicated (read and annihilated in Helsinki) and reconstituted in Washington, in the usual short delay, but in Moscow you are reconstituted after a delay of one year. The question is "does this change the evaluation of the indeterminacy on the W-experience and M-experience?

You can keep your "100%" "probability" evaluation, actually. Because the reversal physics/number's theology will not depend on the choice of any means of computing the indeterminacy, only from its invariance properties. Would you still predict 100% to the Moscow-experience and to the Washington-experience here?



You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To post to this group, send email to everything-list@googlegroups.com.
To unsubscribe from this group, send email to 
For more options, visit this group at 

Reply via email to