From: Eugen Leitl (eugen@leitl.org)
Date: Wed Jun 12 2002 - 05:22:59 MDT
On Wed, 12 Jun 2002, Emlyn O'regan wrote:
> > The key part of the hypothesis we are discussing is that the
> > runs are completely identical. What do you mean by "synched"?
I'm one of the runs. As soon as I realize this, I can lobby the entity I'm
conversing with to remove the constraints (whether trajectory forcing if
it's nondeterministic or identical input if it's deterministic) keeping my
trajectory identical (whether I'm the first or the second run doesn't
matter). The entity may or may not listen. I may or may not have the
chance to do it again later on. But I certainly won't have that choice if
I press the button. So, I won't.
Variations on this involve the number of copies, and how much money you
offer (assuming, there are no other enticements available -- if my runs
are identical, either I'm in solitary confinement, surrounded by zombies,
or other people are in there with me).
> > My scenario is exactly equivalent to a specific Turing Machine
> > (one that is clearly Eugen Leitl, assuming that he has a finite
> > lifespan) being run through twice by the Operating System, or
> > God, or whatever.
> >
> > So I think that they are already "separate people" in your
> > terminology, but I'm not certain.
No, they're not separate people. Yet. They have the potential to become
separate, and diverge rather far apart, if given the chance.
> Hey, actually these are not identical runs. One of the simulations gets the
> offer to push a button. The others don't. Thus there is a relatively
> significant diversion.
They're identical. One of the runs will be terminated. You just don't know
which.
> How would the sims whose runs terminate ever experience that? It would be
You don't experience your termination, even less than a 20 MT blast on top
of your head (assuming, the image execution termination is
synchronous/takes less time than a biological chronon).
> outside their experience. It's not like it'd be a personal tradgedy.
> Interesting.
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:14:45 MST