On 02 Mar 2012, at 22:08, John Clark wrote:

On Thu, Mar 1, 2012 Bruno Marchal <marc...@ulb.ac.be> wrote:

> The question is "If I throw a coin, what is the probability that I see it becoming a flying pig". In front of the UD, that question is not trivial.

In this thought experiment the meaning of the word "I" is not obvious and in fact the entire point of the exercise is supposed to be to make clear exactly what "I" means,

Not at all. "I" (first person) is rather easy, in that situation. If you agree that there is no problem surviving the drinking coffee experience, you have already grasp it.

and yet you throw out the word as if the meaning is already clear. In one sense there is zero probability because if you became a flying pig you would not be Bruno Marchal anymore.

You are right, the question would be silly if it was "what is the probability that I become a flying pig". But the question was

    <<what is the probability that I see it becoming a flying pig".>>

And the "it" refer to the coin, not me. I have to put "I see" before "it" to make clear that I am not asking about the probability that the coin becomes a pig, but only of my seeing of the coin becoming a pig. The probability bears on my future *first person* experience. you have read the sentence above too much quickly.

And in some sense there is zero probability the Helsinki man will be the Moscow man because the Moscow experiences is what transformed the Helsinki man into the Moscow man so that although he may remembers being him he is not the Helsinki man anymore.

But he is still the same guy in the same sense that I am the same guy after drinking a cup of coffee. If you were right, then the probability that *I* see head (or tail) in *any* throwing of the coin would be zero. All probability that I see or feel something after some laps of time would be zero. In particular, the probability that you survive with an artificial brain would be zero, and that would make comp false.

So the answer to the question "If I change what is the probability I will remain the same?" is zero. And that's why I think this first person indeterminacy stuff is just silly.

Not, it is just "if I do that experience, what is the probability that I feel to get that result".

> Comp is just "I can survive with a digital brain". It is about me, my consciousness, my body

Fine, but then how does that square with your comment "Comp makes arithmetic a theory of everything". Consciousness is not everything.

That's my point. Arithmetic is about me, my consciousness, my body, the matter which seems to constitute me, all apparent matter, the laws of physics. Comp makes arithmetic a theory of consciousness including matter's appearance, without ontological matter. Arithmetic is also a theory of numbers to start with, and matter becomes a first person plural appearance, obeying some laws that we can deduce in arithmetic, and then compare with what we see, making the comp hypothesis testable.

> comp makes matter into an appearance in the mind of universal numbers only.

Comp can certainly make a mind that through virtual reality can experience matter that does not in fact exist, but even if the rock the mind feels like he is holding does not exist other matter does in the form of the computer that is simulating the rock, and the mind too.

The point of the reasoning is that this does not work. But you have to follow cautiously the whole reasoning to grasp this by yourself.

You claim you have proven that a computer made of matter is not necessary to do a simulation like this but I'll be damned if I can see where you did this. In Aristotle's metaphysics the potential and the actual are somehow one, but is this really true? I don't know.

Yes. It follows from UDA. It is the part of Aristotle's theory which remains correct in both Plotinus and in the theory of matter of the machine.

> OK. So you see that there is a 1p- indetermination.

I don't even think "1p- indetermination" has a clear meaning except "if you change then you are not the same"; well yes I can see that, it's true but not very profound.

You have still not get the point. It is not "if you change then you are not the same", It is "if you do that experience, what is the probability you feel to get this or that result, when assuming comp so that you agree already that the probability of surviving with a digital brain is 1, despite the big change.

> the question does not bear on where he will be, but on where he will feels to be.

If I receive sense inputs from Washington I will feel like I'm in Washington if I receive sights and sounds from Moscow I will feel like I'm in Moscow. You may ask "why are you the Moscow man and not the Washington man?", and my answer is because I received inputs from Moscow not Washington. So a legitimate question and a proper use of probabilities would be "What is the probability I will receive sights and sound from Moscow but not Washington?". Unlike your question this one is perfectly clear and is well suited for statistical analysis, but I don't see what deep philosophical insights can be gained from it.

It has always been that very question, even above. You have just miss the "it" in the sentence. So you do agree with the first person indeterminacy. It is just a matter of doing all the steps. There should not be any problem up to step 7, where the reversal physics/computer-science occurs. Step 8 is more subtle, but let us first be sure you get the step seven.

> he know that he will be in W and in M, but he knows that whatever he will feel to be, it will be in only one place, because he knows that he will not feel to be in two place at once.

Even that is not a given. This is virtual reality after all,

No. Here I was back on step 3. Not step 7.

it's the point of your dovetail machine, so there is no reason you couldn't have the White House in the middle of the Kremlin and the Washington Monument right next to St. Basil's Cathedral.

In step seven, yes. good. you begin to see that if front of the UD in a robust universe, I have to take into account *all* my consistent extension.

> he is aware that he cannot predict which one among the many "he" he will feel to be.

That is true ONLY if he does not know if he will receive signals from Washington or Moscow, if he knew that, and there is no reason in theory he could not, then he could make such a prediction.


> That is the 1-indeterminacy, which is crucial for the rest of the reasoning.

I know it's crucial, and so if that fails, and it does, then the entire proof falls apart.


Don't misunderstand me, I'm not saying your conclusions about numbers are wrong and in fact my hunch is that they are probably right or close to it, but I don't think you've proved it and I'm certain this 1p indeterminate stuff is a dead end.

From above, I can see you do grasp the point. You have just miss the "it" in the initial question, and that gave you the feeling there is some non sensical stuff there.

> There is no difficulty. Just the discovery of how to explain a objective account of a feeling of subjective indeterminacy in the mechanist framework.

The explanation is not difficult, you never know what's coming next. Forest Gump had a similar explanation that was every bit as deep, "Life is like a box of chocolates...you never know what you're gonna get".

But this simply false. For example we know with certainty that "shit happens". More seriously i know that if I put water on the gas, there is a high probability that I will will the water becoming hot, and a small probability that I will feel the water becoming frozen. I do that experience all morning since 40 years, and that theory works very well. The problem for comp, is that now, I can no more explain it by referring to some real water, real fire, but only by the statistics on all computations in the UD going through my current computational state (in the 3p view). So, if the usual laws on water and boiling are correct, it means that they should be justified by a statistics on computations in the UD (in step seven where we have still a concrete UD running in a primitively physical universe. Then step 8 makes clear we don't need to run the UD, and the primitive physical universe is a red herring.

>> Non-comp may not be contradictory but all the human practitioners of non-comp most certainly are, every single one, no exceptions.

>  Many are, but why all, and why necessarily?

All non-comp fans say that knowing what someone or something does is not enough to determine if it is conscious, you need to know HOW they do what they do; and yet until very recently nobody had the slightest idea how the brain worked and yet they still firmly believed that their fellow human beings were conscious when they acted as if they were, that is to say when they were not sleeping or dead.

Hmm... OK. Most people defending non-comp are not quite rational. Craig exemplifies this very well, by being forced to eventually mention that he *knows* the truth on this, which makes no rational sense.

Even today 99.9% of the human population thinks that how the brain works is so unimportant that they have not bothered to learn the first thing about it and are no more informed on the subject than a medieval alchemist, and yet they still insist that how the brain does what it does is far more important than what it does. This is inconsistent, massively inconsistent.

I can agree. But as a logician, and in the fundamental matter, it remains important to understand that we, the comp people, does not know the truth of comp. That would be the same irrationality than "knowing non-comp". It is just a question of being logically valid.

>> A God needs to be a person, a infinite person perhaps, a superhero of superheroes, but a person.

> Not at all. This is a Abrahamic axiom. There are other theories.

It's not a axiom and it's not a theory it's just the way human beings use the word "God" in the English language.

Since Moses. Probably since before. I like that theory. It is an informal theory, even if it contains fairy tale aspect. let us just be cold and unemotional so that we will be able to compare machines' theology with all the human theologies. But the greeks often discuss the posiible personhood of God like if it was an open problem, and in comp, it is an open problem.

If you want to talk about something that is responsible for our existence but is not a person then that's fine, I do it all the time myself, but when I do so I don't use the word "God" because I believe the purpose of language is communication and thus I believe that a language known only to me is of no use to anyone. There are about 900,000 words in the English language and I'm sure you can find at least one of them that describes Arithmetic better than "God", however as I said before, many people are willing to abandon the idea of God but for reasons I don't understand they are unwilling to abandon the word "God".

Because it is the word the most used to point on the mystery of existence. They are willing to abandon some conception of God, but remains open to use the older general concept, and to get other theories on it. It is theology because it concerns afterlife, the origin of things, the truth we may intuit but not justified, etc.

Note also that I am not saying that GOD = Arithmetical Truth. I am just saying that in the theology of a little Löbian machine like PA, the mathematical (but non arithmetical) notion of Arithmetical Truth (AT) obeys the most common axiom for the notion of God. Mainly, for PA:

- For PA, AT can be obviously said to be responsible (logically) for its existence, and the existence of its consistent extension. - For PA, AT is not nameable (Arithmetical truth is not arithmetical, by Tarski theorem, and PA has no name/description of it, although it can intuit approximation) - PA is in trouble, lose correctness, if it gives a name or a description to it (so there is a sort of notion of blasphemy for machine!).

>> Arithmetic is not a person, super or otherwise.

> Arithmetical truth? How can you be sure.

Do I really need to be sure to conclude that referring to arithmetic as God is unlikely to produce optimal communication fidelity?

Please, use the term "Arithmetical Truth", because "arithmetic" is ambiguous, is it RA, PA, or the set of all arithmetical true proposition? I use "arithmetic" only when I don't care on the nuance, but with respect to god notion, the nuance is capital.

Using God might help the communication, by making a lexicon between Plotinus and what the mystical machine can say about her possible reality and realities. If we don't use the word "god", the fairy tales will continue to drive us for long.

By using the word God, we show respect to our predecessor, and can help the comparison. This includes the discovery that many theology were quite good, despite the politically despairing context of the field.

 > non-plausible does not mean contradictory.

True, but it's the next best thing.

> With different statistical weight, at least for the experience which can be distinguished in the future. If not comp predict white noise for all experience, given that the UD generates them all.

So let's say that every 10^-43 seconds (the Planck Time) my life branches in two (or any finite number), nevertheless when the thing called "I" looks back even after many years have passed he sees a continuous path. There are a enormous number of other things that call themselves John K Clark, some of them very similar to the thing that is writing this Email and others radically different (including one that is now Pope) but all of them see at least some order in their past. It is true that there are a even larger number of paths that have no order whatsoever, no rhyme or reason; and at the end of one of those paths if there were a being looking back he would indeed see nothing but white noise, but there is no being there to look back, because although a little randomness is helpful to Evolution in producing intelligence, random mutation is only half the story, the other half is nonrandom natural selection. So if looking back I saw nothing but white noise I would not be around to look back.

Not bad. With comp, made explicit, once you get the UDA-7, you know that all this can be made mathematically precise, and that the only reason possible for lawfulness of observations will be by weighting the computational histories in the UD, or, equivalently in the tiny sigma_1 part of arithmetic. UDA1-6 already justifies (retrospectively) most of quantum weirdness, for it shows that comp implies indeterminacy (like QM MW), non locality (like QM), and even non-cloning of matter (like QM, that's an exercise once you get step-seven). Unlike Everett QM, we must justifies BOTH the Born rule AND the SWE, but the computer science math suggest that nothing prevents this to be done, and I have extracted already a tiny part, and show precisely how to continue. The only weakness is that it leads to complex math. But the goal was to make the mind body problem precise, and it would have been astonishing that it would lead to something simple.



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