Re: Simplicity, the infinite and the everything (42x)

From: Quentin Anciaux <>
Date: Wed, 13 Aug 2008 01:03:50 +0200


2008/8/13 1Z <>:
> On 12 Aug, 22:50, "Quentin Anciaux" <> wrote:
>> Hi,
>> 2008/8/12 1Z <>:
>> > --- In, "Quentin Anciaux"
>> > <> wrote:
>> >> 1) Why 1 is more than 0 and simpler than n ?
>> >> 'Entia non sunt multiplicanda praeter necessitatem'... It follows by
>> >> looking at it in a first sight that it would means the one universe
>> >> hypothesis is simpler than MW. Yes, one universe involves many less
>> >> than MW (either there exists a finite number of other universes or
>> >> infinitely many)... Then by O'R we should take the one universe
>> >> hypothesis as simpler because requiring less universes (in this case
>> >> 1). But it's an ill way of understanding O'R... It should be
>> >> understood as saying something about the premises, the axioms...
>> > Why ? Occam talked about entities. And does MMW really
>> > minimise the number of axioms? Doesn't it need the axiom
>> > that nothing non-mathematical exists?
>> And doesn't the one universe has a 'real switch' axiom that only
>> something exists ?
> Yes, but One Universe (or at least, non-MMW) methodology does
> not claim or pretend or wish to have 0 axioms. I aims for an
> ontologically
> parsimonious explanation that matches the evidence.

Yes so ball at the centre... these axioms are equally valable...
except that the finite number of universes hypothesis has to explain
why that number (be it 1 or 42).
Could you explain more precisely what is parsimonious for you ?

>> >> One
>> >> shouldn't add an axiom unnecessarily. And in this case none of one
>> >> univers, 42 universes or infinitely many hypothesis are simpler
>> >> relative to each other...
>> > That doesn't follow on either interpretation of O's R. It is
>> > obviously false on the "entity" interpretation. On the "axiom"
>> > interpretation, if there are exactly 42 universes, and there is no
>> > strict necessity to the fact then, the "the number of universes is 42"
>> > is an axiom.
>> Such as "the number of universes is one".
> One Universe (or at least, non-MMW) methodology does
> not claim or pretend or wish to have 0 axioms.

Well so ?

>>Besides I find very
>> problematic the unicity.
> Then you had better say what the problem is.

Why one ? why not two ? if one why stay one ? There is a single
universe and a precise one and *only this one* and we are actually in
it, wow lucky.

>> >>and O'R could not help you choose or if it
>> >> could help for something would be to choose the 0 univers
>> >> hypothesis... well 0 < 1 << oo and this for all values of 0 even big
>> >> ones :)
>> > O's R is of course about the simplest theory that fist the evidence.
>> Well then infinitely many is simpler (does not have to bother to
>> explain why one and only one for ever and ever).
> it is not simpler on the "entity" version of O's R, and it does not
> fit the evidence because of the WR problem.

Yes but I see 'real switch' problem as equally problematic in front of
the WR problem.

>> >> 2) Why turing emulability of the mind entails first person
>> >> undeterminacy and/or MW ?
>> >> Because if you're a computation then you're not dependant on the
>> >> substrate of the computation... but only to the computation itself.
>> > I would be directly dependent on the computation but only indirectly
>> > on the substrate.
>> You have no dependancy on the substrate (even indirectly), the
>> substrate can change and the you computation can't notice it (so the
>> direct substrate supporting the you computation may be an
>> emulated.substrate running on an emulated substrate running on your
>> level 0, still the you computation doesn't care.
> The computation needs some sort of substrate.

*Any* substrate that can be use for doing a computation. Is a program
running in a simulated processor can know that the substrate (imagine
it has sensor to give information about it) is simulated if the
simulated processor gives out exactly what it should ? no... so the
substrate is nothing.

>> > A
>> >> computation is substrate independant.
>> >> Well you'd say it may be substrate independent but still it needs a
>> >> substrate to 'exist'. Ok let's accept that, but let's return on the
>> >> mind and on the hypothesis that the mind is a computation and the
>> >> brain the substrate on which it is run. As a computation is substrate
>> >> independant then what follows is if the mind is a computation it can
>> >> be run on other computational substrate for example on a... computer
>> >> for example. And 'the mind' wouldn't be able to tell if it is run on a
>> >> brain or on a computer. By our hypothesis the mind is a computation,
>> >> and a computation is
>> > You have switched context here from what a computational process is
>> > ontologically dependent on, to what the abstract description of an
>> > algorithm is theoretically dependent n
>> No you devise this in 2 parts, I think only the abstract world is
>> ontologically primary.
> That is your conclusions. You cannot assume it in order to
> argue for it.

I do not assume them.

>> >>dependant only on it's state and transition rule,
>> >> if the same input is given to the same algorithm it will yield the
>> >> same result so seeing a brain is of no help because you would see a
>> >> brain even run somewhere else if the same input is given.
>> >> So why this entails first person undeterminacy and/or MW ?
>> > let's
>> >> assume we could replicate the computation of your mind (I have assume
>> >> by hypothesis that it is a computation, so replication can be done,
>> >> even if currently we don't have a clue and even we don't know if the
>> >> mind is a computation... but here I assume it just for the argument to
>> >> see what it entails) then I could execute the 'you' computation on a
>> >> computer then if I can, I can also run the 'you' computation not only
>> >> on one computer but on many computers. Ok so now I have at least two
>> >> computers running the same mind (computation)... I switch off one
>> >> computer, the mind die ? hell no, by our hypothesis mind is
>> >> computation and the computation is still running on the other
>> >> computer. So from the point of view of the mind unplugging one of the
>> >> two computers didn't change a thing. Now I'm a real serail killer I
>> >> switch of the last computer running the computation/mind... so now the
>> >> mind die now ?? Let's say I've done a program dump before stopping the
>> >> last computer and I decide 5 years later to rerun the computation from
>> >> this save point and on. Wasn't the mind dead ? If it is and mind ==
>> >> the computation, how can I have the ability to run the computation
>> >> without it being the mind ? It means also that if you're a computation
>> >> you can't know at which 'level' you're run (if you're run on a VM
>> >> running in a VM running in a VM or a non emulated substrate). So if
>> >> mind is a computation to make correct prediction about the next state
>> >> you must take all computation having the same state into account.
>> > No. If I have no reason to believe there multiple implementations,
>> > there s in no indeterminacy between them. I can't be certain that
>> > there ae no multiple implementations, but I can't be certain that
>> > I am not being dreamt by a butterfly either. The problem with
>> > sceptical arguments is that they confuse "you cannot be certain
>> > that not-X" with "you can be reasonably sure that X". I am entitled
>> > to reject the Yes Doctor hypotheses along with the Chuang Tze
>> > hypothesis even if I cannot be certain.
>> If you're a computation then you are duplicable as a computation,
> DuplicABLE is not enough. I might be dreamABLE by the butterfly as
> well. If I am not duplicATED there is no indeterminacy. If there
> is no serious evidence of duplication, there is no serious problem
> of indeterminacy -- it is just speculation like the butterfly story.

If mind is a computation it's a fact that conscious computation will
be run. If mind is not a computation on the other hand well yes mind
wouldn't be duplicable so asking for potential duplicability in these
conditions is meaningless.

>> this
>> could be seen only an engineering problem not a technical one (if
>> you're a computation of course). Then either the mind is not a
>> computation so we could not duplicate it (not as a computation
>> actually) or it is and then it will be possible.
> All sorts of things are possible that I don't worry about. Why should
> I?

The question is about mind is or is not a computation and what it
entails. You are saying mind is a computation but are rejecting
multiple runs on the ground that there won't be any ever... Either you
should dismiss the mind computation hypothesis or dismiss your
susbtance real switch theory, they're not compatible.

>>Then either the you
>> computation is run once and will never and ever be rerun (on any
>> substrate able to support this computation, even a sheet of paper :) )
>> or it will be run more than once, then the rest follows.
> There is not a scrap of evidence for the more-than-once option.
>>The only way
>> to be sure not to be rerun is that the mind is not a computation... It
>> can't be a computation and not imply 1st person indeterminacy.
> You're doing it again...putting forward "you can't be sure of not-X"
> as if it means "you can be sure of X".

No I can be sure that if mind is a computation it is a computation in
pure term and thus totally independent of the substrate.

>> > Even
>> >> using the 'real switch' theory a mind could be run on different 'real'
>> >> (composed of substance) substrate.
>> > "Could" does not cut it. If it is not reasonably
>> > plausible that there are multiple implementations, there is no
>> > indeterminacy worth worrying about.
>> Could... as I say the "one" univers we see is big and has plenty of time ahead.
> The universe we see is nowhere near as big as the Tegmark-Marechal
> Platonia.
> Therefore you cannot possibly lever Mathematical Many Worlds out
> of Physical Many Worlds.

nowhere as big... you've seen the entire universe or only what's in
your light cone ?

>> >>.. and the mind will *have* to take
>> >> into account these runs on real substrate to make correct prediction.
>> >> And unplugging one real substrate run will not kill the mind,
>> >> unplugging them all also. The only way would be to not only unplug
>> >> them all but to garantee that it wil *never and ever* be run *again*
>> >> (even only one).
>> > In reality, no-one has ever come up with a perfect duplicate of even
>> > one mind. You just don't have your multiple implementations.
>> Yes, but that's not the point, the point is assuming turing
>> emulability of the mind/the mind is a computation hypothesis implies
>> 1st person indeterminacy and hence many worlds/dreams.
> False. The TE alone does not imply indeterminacy. To obtain actual
> indeterminacy,
> you need to get actual -- not just theoretical--implementations from
> somewhere. You need the additional hypothesis of Platonism, or
> something else.

Mind emulability implies indeterminacy (even for only two runs).

>> So assuming
>> mind is computation we will have multiple implementations,
> "Will" is nothing but a guess.
>>*if* mind
>> is a computation the problem lies in an engineering problem not a
>> theoretical one.
>> >> If I'm run on another computational substrate than
>> >> my brain, If someone pull the plug, I die ?
>> Still don't think so or I'm can't be run on another substrate (ie: not
>> a computation) in the first place.
> That's as much faith as any religion. You do not know that someone
> will
> go to the trouble of duplicating you at some point in the future.

Yes so the next you is a relative state of your current computational
state, your current state + transition rule + next state is part of an
infinite number of computations. As Bruno is saying you are the trace
of all those merging computations.

Quentin Anciaux
All those moments will be lost in time, like tears in rain.
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to
To unsubscribe from this group, send email to
For more options, visit this group at
Received on Tue Aug 12 2008 - 19:04:05 PDT

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:15 PST