Quentin Anciaux writes:
> What I understand from that is as if you could influence probabilty, as if 
> knowing something or acting in some way will change your "future" Hal by 
> having him "good moments"... But if at every choice, every results exists 
> (whatever the measures of each one).. Some Hal are always garanteed to have a 
> "bad' experience and some other garanteed to have a "good" experience, 
> whatever you knew before the choice, whatever the measure. So my question is 
> how could you think acting "good" change anything in this matter ? And also 
> what is "good" and what is not "good" and how you classify these ?

I'm not sure I understand exactly what you are getting at, but here are
a couple of ideas that may help improve our communication.

One issue with a multiverse concept is that since "everything will happen"
then it seems that we do not really make choices.  We don't have free
will.  When we choose between A and B, even if we take a long time to make
the choice, there will still be universes where we choose A and universes
where we choose B.  So it seems that we did not make a choice after all.

To this I make two responses.  The first is that even though
everything will happen, not everything happens with equal measure,
or equal probability.  For a multiverse model to make sense we must
(in my opinion) include the notion that some universes are more likely
and have greater measure than others.  Otherwise we cannot explain why
natural laws continue to work and our universe does not degenerate all
around us into chaos.

Given that different universes in the multiverse have different measure,
this puts a different light on choice.  Although it may be true that
we will make both choices in some universes, by our process of careful
thought and consideration, we do make a difference.  We change the
measure of the possible universes.

If we just flip a coin to make the choice, then both outcomes will
happen with equal probability and both universes will have equal measure.
But if we think about it carefully and then decide to do A, that gives
more measure to the universe where A happens.  If we consider all versions
of ourselves, most of them will make that same decision, and most measure
will go into the universes where that is the choice.

So the nature of choice in a multiverse is that it is a matter of changing
the measure of the universes where the consequences happen.

The second point related to this is that some people may object that even
the appearance of making a choice is an illusion, because our thoughts
are determined by events and causes outside of ourselves.  This is an old
argument and is not specific to multiverse models.  Suffice it to say
that there is an extensive philosophical literature, and in particular
the doctrine of "compatibilism" describes how we can think of making
meaningful choices even though we may live in a deterministic universe.

Hopefully these ideas will help you see how it does make sense to make
choices and take actions even in a multiverse.  This is not specific
to my thought experiment, it relates to everyday, ordinary action.
Every breath we take is a choice to live.  Even in a multiverse it makes
sense to think of our actions in this way.

As far as your final question about what I meant by "good" in my
experiment, it's not very important what the details are.  I just
want to distinguish relatively pleasant events from those which are
not so pleasant.  The main point was to shed light on the question of
how we can reconcile two models of experience: a linear model where
we pass through life in a straight line from birth to death; or a
random-sampling model where every moment of every person's life exists
somewhat independently, and we could imagine a probability distribution
over those observer-moments which determines how likely they are to
be experienced.  My thought experiment was an attempt to describe a
situation where the two views would seem to give different answers
about what we would do, to help us understand how they differ.

Hal

Reply via email to