Re: Worlds do fuse (please delete the previous mail)

From: Gilles HENRI <Gilles.Henri.domain.name.hidden>
Date: Mon, 14 Jun 1999 14:15:23 +0200

Sorry, the last mail was incomplete

>
>Question remaining is the problem of identity fusing. How is this done?
>Identical minds are actually identical consciousness? By how much should a
>mind be different to generate a new consciousness? Is consciousness discrete
>or is it a continuum such that we share the same continuum?
>
>George

I think that it is indeed the most difficult point in MWI. In fact there
are not many worlds. There is a single, huge Universe containing many
apparent classical worlds. But what defines the "borders" of a classical
world? An apparent world is profoundly linked to the consciousness we have
from it, so this question is quite the same as George's one : what defines
a conscious state?
 We could think of some set of projectors defining "quasi -classical"
states and think that the consciousness must be attached to some peculiar
classical states, but defining properly what is a quasi-classical state is
itself very difficult. It can only be defined "grossly speaking", so we
must admit that our own state is not precisely defined...I wonder if the
"pure computation" hypothesis would answer this problem, since it has also
to define which computations would correspond to a given conscious state
(and more generraly to any given quasi-classical world).

If an electron passes through 2 slits, it is not correct to consider that
the two possible trajectories correspond to two different macroscopic
worlds, because they are still strongly interacting : rather we will
consider it as a single quasi-classical world containing some quantum
delocalized particle, for which we can apply QM rules. Only very weakly
interacting states could be considered as two different worlds.

This is also strongly linked to the question of past and future, since
these notions make sense only for quasi classical states. Strictly
speaking, the number of past AND future states is infinite, because there
is a non zero probability of evolving from any quasi-classical
configurations to any other one. However, very tiny probabilities (i.e.
between very different macroscopic states) correspond also to apparently
temporally non-connected worlds. So if we restrict to temporally connected
worlds (by definition states for which the transition probability
calculated by QM is O(1)), the number of possible past and future states is
much lower. Although during quantum measurement a single quasi-classical
state can give rise to multiple futures, it is much less obvious that a
quasi classical state can arise from several CLASSICALLY different past
states : if past states are macroscopically different, the transitions
erasing completely their differences are very improbable, and so do not
correspond to classical temporal transitions. So I think that a classical
world can have only one classical past, and many classical futures. I
assume that it is related somehow to the second principle, but I haven't
found exactly how yet.

Gilles
Received on Mon Jun 14 1999 - 05:19:14 PDT