Re: Simulation argument in the NY Times

From: Vladimir Nesov (robotact@mail.ru)
Date: Fri Aug 17 2007 - 03:50:52 MDT


Friday, August 17, 2007, Norman Noman wrote:

NN> Strictly speaking, to "simulate" entails REPRESENTING certain key
NN> characteristics or behaviours of a selected system. Representation is an
NN> intentional act, and only intelligent entities have intentions. However, if
NN> due to profound coincidence a computer running simcity popped out of nowhere
NN> through quantum uncertainty, most people would still consider that a
NN> simulation, and the unlikelihood of this sort of event is what I was talking
NN> about.

It's a question of being prepared. Say, you are being copied, so that
1000 copies of you will be sent to labor camps and 1 copy of you is
given a substantial money compensation. One moment you are sitting in
a copying device, next moment there are 1001 of you. Which of them do
you expect yourself to become? With what probability?

This question is not correct: you will become each one of them.
Expectations before the procedure can only prepare you to operate
more effectively is _some_ of future situations. This is a tradeoff:
you can't have general expectations, so in some of the futures you
will be less prepared than in others. Choice of futures to prepare for
is arbitrary, such as correlation to certain measure of this or that
outcome. This is a kind of remnant of evolution, which tried to infect
as much future branches as possible with given DNA.

-- 
 Vladimir Nesov                            mailto:robotact@mail.ru


This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:58 MDT