Re: valuable errors

From: Gilles HENRI <>
Date: Fri, 16 Apr 1999 21:03:47 +0200

>Jacques M Mallah wrote (about self-multiplication):
>> The situation you described is completely deterministic, much like
>>the MWI of QM.
>OK, I completely agree with that. At least if the description of
>the situation is conceived as a third person type of description.
>> For all practical purposes, a person who is copied should expect
>>their future selves to be effectively randomly chosen.
>I am glad you agree with that. This is really the first person
>indeterminism which appears naturally with comp. (And with MWI).
>You note that there is no inconsistency between first person
>indeterminism and third person determinism.
>Suppose now some person, X0, is copied in (exactly) 1O instances. She
>expect their future selves to be effectively randomly chosen, as you say,
>among the 10 instances: X1, ... X10. (or 11 if the original is not
>Do you agree that, whatever the means used to quantify the indeterminism,
>the immediate expectation of X will not depend:
> a) neither of the "real"/virtual nature of the reconstitution,
> b) nor of the absence or presence of finite delay of reconstitution?
>Thanks for answering so far.
>PS You say also:
>> If you want to talk about what is actually going on though, I
>>don't even accept that 'individual identity' carries over from one time
>>step of a computation to the next. It's just that the future self or
>>selves are sufficiently similar to the current self to motivate an
>>interest in his (or their) well being.
>I agree too, but I am not trying to know what is actually going on.
>I am interested in the practical first person expectations.
>I use the notion of "survival" in the thought experiment for
>escaping the need to define precisely "individual identity".

the question is important because Bruno's theory relies entirely on the
question whether you survive a teletransport, i.e. if your copy is really
"you". I think you cannot escape to precise what is the "individual
identity" by a semantic artefact.

In my sense, as I already explained it is partly a subjective question,
because there is no objective definition of identity. However we would try
to agree on a definition that is both paradox-free and that allow to
recover the "normal" identity that seems to exist in a classical world.
Bruno's hypothesis is that we are perfectly (FAPP) reproducible at some
finite level and that any copy at this level has the same right to be
called "us"(COMP). It is only because we have currently no way to make such
copy that we are unique.

I have two criticms to this point of view :

1)COMP is probably false. we are made in such a way that I don't believe we
can be "reproduced" by another physical structure. Our behaviour seems
superficially close to that of a computer, but in many aspects it is not,
because we included our brain) are also (and first) a chemical system that
reacts "continuously" and "analogically" to our environment, for example by
taking alcool and other drugs, by suffering from diseases from toxins
secreted by microorganisms, and more generally reacting to our own hormons
that govern our emotions (all things that ARE important in our life). If
you want to incorporate these things, you have to simulate a human down to
its molecular structure, but as I argued it is exactly the opposite what a
computer is. A computer has a number of physical degrees of freedom
(physical entropy) enormously greater than the number of its computational
degrees of freedom (memory and processor size); that allows to reproduce
the same computational complexity with many different material structures.
So it is clear that if you want to simulate a physical system (down to
detailed molecular structure) with a computer, you will need a computer
huger than this system. But then this computer cannot behave PHYSICALLY
like this system. The only possibility is to built a "molecular" computer
that has exactly the same PHYSICAL behaviour than your system, that is in
fact an exact PHYSICAL copy of you (usually what SF authors assume!). What
I argue is that in fact we don't know how built a physical copy (down to
quantum level) of any MACROSCOPIC system (larger than one photon!) , and it
is very probably fundamentally impossible. So you can at most hope to built
a reasonably "fake" copy of yourself, but that would sooner or later
separate from your own behaviour even if it is in the same environment. I
would find it difficult to consider this copy as myself...

2) Even if COMP were true the fact that identity is entirely determined by
the "functionality" is discutable. In most cases, it is determined by the
physical spatio-temporal continuity. Here are several examples:
 - if you built two "identical" computers, they won't be considered as
"really" identical even if they do exactly the same thing. For instance if
you happen to break your computer accidentally, you will not be allowed to
exchange it with that of your neighboor. In case of a dispute with him,
people will try to look at some detail (for example the serial number),
which have nothing to do with the functionality, but which will decide
which computer has been "physically" (continuously) yours before the crash.
Same with any usual object (car, etc...)
- inversely the broken computer is always your computer, even if it has no
more functionality.
- same with human beings. If a twin die for example just after a birth,
his(her) corpse is associated to his identity. He will not be considered as
surviving in the other living twin (in our culture at least...), although
the latter is much closer "functionnaly" to his former living state.
- more generally we can associate an identity to temporally continuous
structures even if they suffer profound changes. We can say "scientists
have found the skeleton of a dinosaur" even if it has no dinosaur-like
So there is no reason to consider that "you" survive in a copy. Of course
you can assume it is true, just as you MAY believe you are the
reincarnation of an ancient Egyptian prostitute (yes, some famous people do
believe that!). But this kind of belief cannot learn anything on the real
nature of the world!

Some of you treat exactly the potential duplication in our real world like
the MWI. But it is quite different. The "many worlds" in MWI are defined by
the fact that they are almost not interacting. We "duplicate" in the MWI
sense exactly when and because our "copies" cease to interact. This insure
the unicity of each copy in each "macroscopic world". MWI does not allow
several copies to know each other. Classical probability does not apply
(like in any form of QM) because in fact you are here AND elsewhere.
Conversely the indiduality is conserved in MWI because some worlds are
connected by the appearance of a temporal evolution, which DEFINE who is
connected to whom. Each copy of each world is connected by "classical
histories" to its past(s) and future(s), which insure the spatio-temporal
continuity which would not exist in a hypothetical teleportation (which
remains up to now a dream!)

Received on Fri Apr 16 1999 - 12:05:49 PDT

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:06 PST