Re: ESSAY: How to deter a rogue AI by using your first-mover advantage

From: Vladimir Nesov (robotact@mail.ru)
Date: Sun Aug 26 2007 - 08:25:23 MDT


Sunday, August 26, 2007, Stathis Papaioannou wrote:

SP> Of course, the problem with Pascal's
SP> wager is that there are so many actual and possible religions that you
SP> may as well ignore all of them, unless you find good evidence that a
SP> particular one is true - which translates to evidence that you are in
SP> fact in a simulation.

It wouldn't help either: what matters is observable laws of the universe. If it's
a simulation under control of complex system which is outside the universe
(FAI-running-the-simulation telling you-in-simulation what it is, what to do, enforcing
its will, etc.), you have no way to reliably model that control
system. With our-universe-like universe, basic physical laws are
relatively simple and stupid. You can dissect any process down to the
simpliest physical processes, and you can predict outcomes of complex processes
based on this underlying simplicity. You can reasonably expect
simple models based on experience to extrapolate to the future. With
complex control system that is outside your reach it's impossible:
future can be _independent_ of whole of past history, causation
machinery is not observable by inhabitants of simulation. For this
same reason you can't know whether you are in one simulation or in a
completly different simulation, independently of kind of miracles and
'inside-info' controlling agency claims to be presenting to you.

-- 
 Vladimir Nesov                            mailto:robotact@mail.ru


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