[ExI] Uploading and selfhood

Stathis Papaioannou stathisp at gmail.com
Tue Apr 1 10:57:28 UTC 2008


On 01/04/2008, ben <benboc at lineone.net> wrote:

>  I don't see why people who make statements like this think they are
>  saying something reasonable. "If you get copied EXACTLY, then one of the
>  copies won't mind being killed" !?!?!
>
>  Why on earth would anyone think this?
>
>  I do think that an identical copy of my mind would be me, ('a me', if
>  you like), and i'm sure that none of the me's would be happy (or
>  relaxed, or indifferent) to be killed.
>
>  I remember reading a short story in which people start committing
>  suicide after the discovery that the multiple-worlds interpretation is
>  in fact true, and everyone has an infinite number of versions of them in
>  an infinite number of universes. I also remember thinking that this
>  story _makes no sense_. Would YOU commit suicide just because you knew
>  for a fact that there was another you somewhere? It's just a daft idea.

Well, I'll put my hand up to join the daft. If there were two copies
of me, A and B, *running in perfect lockstep*, let's say in
deterministic parallel computations, then I don't see what possible
difference it could make to me (that is, either one of the me's) if
one of the copies suddenly stopped.

Suppose I am copy A. At time T, the machine running A is turned off
while the machine running copy B continues. If you think that I, copy
A, don't survive the procedure consider this alternative scenario.
There is only one copy of me running on a particular machine. At time
T, the computation is saved to disk and the machine turned off. The
saved data is then loaded into another machine running the same
program, which therefore starts off at the point where the first
machine was stopped. Will I survive this procedure or will I die where
the computation is stopped and saved? Because this situation is
exactly equivalent to copy A stopping while copy B continues.





-- 
Stathis Papaioannou



More information about the extropy-chat mailing list