Re: Many Worlds and Oracles

From: Hal Finney <hal.domain.name.hidden>
Date: Sat, 4 Jan 2003 11:32:51 -0800

Saibal Mitra writes:
> I am not convinced that the MWI implies that quantum suicide should work. A
> hidden assumption proponents of quantum suicide make is that once you are in
> a certain branch all the other possible branches are off limits to you. You
> will forever move on in that branch. I reject this. I could also survive
> (with memory loss) in another branch in which I never performed the
> experiment in the first place. Normally this assumption doesn't affect
> probabilities, except in cases were the observer doesn't survive. I would
> say that, in principle, a theory of everything would allow you to compute
> the probability of experiencing a ''personal history'' x, by integrating
> over all possible histories X of the multiverse that are consistent with x.
> Within this framework there is no room for quantum suicide to work.

This has been discussed at length in the past on this list, without much
resolution. You might want to clarify what you mean by quantum suicide
"working". What do you hope to accomplish via QS? What effect will it
have on your subjective perceptions?

I think we all agree with the objective consequences of QS. Someone kills
themself in some branch of the multiverse. They are removed from that
branch but continue to live on in others. The disagreements begin when
we consider the subjective effects.

The most limited view, I think, is that QS will subjectively fail.
That is, if you attempt to kill yourself in some clean way, the machine
will work in some universes and fail in others, but you will only notice
the result in those universes (or branches of the MWI) where it fails.

In single-universe models, we would predict that a highly reliable
QS machine will simply terminate your existence. You won't perceive
anything, after it goes off. In multiverse models, a QS believer would
predict that you will continue to exist even after the machine goes
off. Your measure will be reduced, but your existence will continue.
This is the most limited and restricted prediction with regard to QS in
a multiverse.

(With some assumptions, you could get the same prediction of continued
existence in a single-universe model. If we assume that the universe
is infinite in extent so there are multiple copies of you, or that some
future civilization will simulate your mind perfectly, then this could be
a mechanism for your existence to continue subjectively even after the QS
machine goes off. But in a finite and "small" single-universe model with
limited future computing capacity, then a QS death is final and complete.)

A more ambitious version of QS attempts to change and hopefully improve
your circumstances by pruning those branches where bad things happen.
By doing this, your total measure is reduced, but your average happiness
(averaged over all instances of yourself) is increased. Because your
total measure is reduced, your total happiness is decreased (unless
you only use QS to eliminate branches where your happiness is negative,
that is, where you are better off dead). So there is a tradeoff between
increasing your average happiness while decreasing your total happiness.
I believe it is a matter of personal preference which one is better, so I
think on this basis that QS is right for some people and wrong for others.

Another ambitious flavor of QS predicts that we can distinguish multiverse
from single-universe realities by attempting QS. It is even possible
that all versions of death fail just like QS would, hence that everyone
who has ever lived has experienced subjective immortality, if multiverse
theories are true. By this reasoning, each person will eventually become
subjectively convinced of multiverse theories (if they deduce that this
is the explanation for their bizarrely extended lifespans).

Hal Finney
Received on Sat Jan 04 2003 - 14:34:21 PST