From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Sat Apr 05 2003 - 15:09:27 MST
When I teleport, I want to anticipate a 100% subjective probability of
being where I'm going, not a 50% subjective probability of being where I'm
going. I therefore want to be moved, not copied.
When I teleport, I prefer a "destructive" move to a nondestructive copy
for the same reason that, given many-worlds theory, I'd rather have a
million dollars in cash than buy a lottery ticket based on quantum-random
information. There may be a 100% probability that *at least one of my
branches* wins the lottery, but I want to anticipate a 100% *subjective*
probability of having the money. Similarly, when I teleport, I prefer a
"destructive" move to a nondestructive copy because I want to anticipate a
100% probability of the "reward", in this case, being in the place I want
to be.
I don't really regard the move as "destructive" because if a pattern goes
from time T to time T+1 through a series of causal interactions that
involve interim storage on a hard drive, that's no more and no less
continuity of pattern than the physical state of my body at time T+1 being
determined by the unitary evolution of quantum mechanics from my physical
state at time T. It's *all* quantum mechanics. The hard drive is quantum
mechanics too. Does it matter how many interactions there are? What
matters is just that you have a set of conditional probabilities which,
applied in sequence, arrive at a developed pattern P2 for time T+1 which
echoes the history of pattern P1 at time T.
It certainly doesn't matter "which" particles you use. You can't possibly
understand, say, Bose-Einstein statistics, and not realize that. Having
photon A at position X and photon B at position Y is not just a state
*indistinguishable* from B at X and A at Y, it is a state *identical* to B
at X and A at Y, the *same state*, *exactly*, which is why we add the two
complex probability amplitudes for ending up in that state.
-- Eliezer S. Yudkowsky http://singinst.org/ Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Sat Apr 05 2003 - 15:21:23 MST