[extropy-chat] Probability of identity - solution?

Russell Wallace russell.wallace at gmail.com
Fri Oct 13 06:03:36 UTC 2006

On 10/13/06, Lee Corbin <lcorbin at rawbw.com> wrote:
> Earlier I wrote that *probability* doesn't provide a very good way of
> addressing
> these problems. I'll note an exception later, however. Another reason that
> probability is inappropriate is that probabilities sum to 1.  In these
> cases, "your"
> experiences sum to 1000, not 1.

But the probabilities of "myself" ending up "being" one of those 1000 do sum
to 1, so I'm not sure that's relevant. After all "myselves" across the
Tegmark multiverse sum to infinity, but my probability of observing a
flipped coin ending up heads is .5 just as standard theory predicts.

What about my objection?  It was (perhaps not well-described) along these
> lines:  Suppose that the A/B split occurs after .0001 seconds, and the
> subsequent splits to B all occur in the next .0009 seconds.  Then all the
> forking is over in a millisecond.  In that case would it not seem peculiar
> to suppose that the single copy experiencing A was as representative
> of you as the 999?   We must banish the notion that there is a soul and
> it's being
> divided into ever smaller pieces!

Sorry, that was part of what I meant by "what you say makes sense" - I agree
that is a problem with the causal logic 50/50 prediction, and indeed as the
time becomes short compared to the timescale of conscious thought, intuition
becomes less sure of the 50/50 result.

The exception I spoke of before, where probability *must* rear its head
> is for *planning* purposes.  Suppose that the 999 of you will be copied
> into a location several hundred feet under water, and only one of you
> copied into a location at STP. Then you should walk around all day long,
> even while driving to work and at the theatre, in your scuba gear. I
> explore
> the feelings that one would have while taking such steps in my old story
> http://www.leecorbin.com/PitAndDuplicate.html , a link I've posted here
> many times.

Sure. The inconvenience of such is less than 1/1000 of the cost of losing
one's life.

Well, I can't go along with that.  It's really not the case upon the usual
> meaning of words, to say that Smith has just died when Smith looks
> and sounds just the way he did ten minutes ago. We really are concerned
> with *survival*, and so we have to reject philosophical stances that
> introduce concepts that wouldn't be useful in any imagined situation.
> (At least in any situation I've ever been able to imagine.)

Sure, but the usual meaning of words breaks down here. An outside observer
nonetheless would say there ends up being 1 A-Smith and 999 B-Smiths. And
that's a valid viewpoint. Smith before the first copy saying "its 50/50 I'll
be A" is also a valid viewpoint.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20061013/528d9c0a/attachment.html>

More information about the extropy-chat mailing list