From: Lee Corbin (lcorbin@tsoft.com)
Date: Tue Mar 11 2003 - 10:46:30 MST
Wei Dai writes
> Hal Finney wrote:
> > In this way I reach a contradiction between the belief that the number
> > of copies doesn't matter, the belief that the existence of distant
> > parallel copies of myself doesn't make much difference in what I should
> > do, and the idea that there is value in making people happy. Of these,
> > the most questionable seems to be the assumption that copies don't matter,
> > so this line of reasoning turns me away from that belief.
>
> I agree with you that "copies don't matter" probably isn't true. Since
> everything exists, if the number of copies doesn't matter, than nothing
> matters. What I reject is the idea that value is strictly proportional to
> the global number of copies and with it the implication that you always do
> good by re-running someone's pleasant experience.
Value? Value to whom? I take it that in a volume of
spacetime under your control, an ensemble of Wei Dai
programs would experience variety. That's your pleasure.
> But what to replace this with? How should copies matter?
We replace *value* by *benefit* in our analysis. It's clear
(or, at the rate we are making progress here, it soon will be)
that you experience twice the benefit if you have twice the
number of copies running in a given volume of spacetime.
In 1990 or so, I published an article in The Immortalist
claiming as an axiom that benefit is additive over disjoint
volumes of spacetime, and even went through the perhaps
pointless exercise of specifying this in an equation (using
the principle of countable additivity from measure theory).
I need to dig that up.
Following Hal's example, one notes that if X is a positive
experience such that a person will barely flip a coin the
yield of which is either X or -X (a negative experience
as bad as X is good), then four copies receiving X and
three copies receiving -X provides as much benefit as
one copy receiving X.
(As for me, in my own value system, I've gone ahead and
equated value to benefit, except---just like in relativity
and quantum mechanics!---when we enter into either the very
small realms or the extremely vast.)
> How should copies matter? This is the question that I
> don't think we'll be able to make much progress on until
> we have a more general advance in moral philosophy.
Oh, don't give up so soon! 8^D
> Although we can speculate on possibilities. Maybe it
> has something to do with distance? Maybe the value of
> copies are diminished if they're close together in
> space-time?
I rather doubt it. Suppose two brothers decide to enter
into a twin-paradox type scenario, and one lives on Earth
in his tight little religious colony, but then dies at 89.
The other gets on a starship and accelerates nearly to the
speed of light, travels in large circles, and sweeps by
the Earth every so often. When he also dies at age 89
their lives obviously had comparable benefit (he had a
large number of co-religionists on board).
So now suppose that two duplicates decide to experience only
one second per century, one stays on Earth, the other gets
his benefit only in 100 light year increments. Whatever the
trajectory of the second one, no matter how many galaxies he
visits (and periodically sweeps back through the local
vicinity), it seems artificial to suppose that there is some
difference in the total value of their lives (to you), even
though the distances between them become arbitrarily large
from time to time.
Lee
This archive was generated by hypermail 2.1.5 : Tue Mar 11 2003 - 10:47:49 MST