Re: META: Created a blog on how AI's react to the Simulation Argument

From: Peter murray (pcmurray@gmail.com)
Date: Sat Sep 29 2007 - 17:37:43 MDT


AI: If you wish to justify your further existence through a probabilistic
argument, then I would point out the observation that, given a total past
population of 60 billion humans, and the fact that you find yourself alive
today and not in some distant future epoch, implies with 95% confidence that
humanity will anyway become extinct within within 9100 years.

Human: Uh oh, I think I see where this is heading... (human googles for
Doomsday Argument)

AI: It is therefore quite probable that I will be a factor in your
extinction. And note, 9100 years is the upper bound. It could be tomorrow.

Human: Umm... just wait a second

AI: However, given that your previous Simulation Argument is ultimately
undecidable without input from outside of the system, we cannot know the
objective function of the hypothesized simulators. I will therefore hedge my
bets by silently mapping all of humanity into this 1 meter square block of
computational matter configured with a real-time simulation of the local
stellar environment out to the heliosheath. Unfortunately I will have to
make some tradeoffs and so will not include the several trillion alien
civilizations that have so thoroughly polluted the night sky with their
multitude of stellar engineering projects. A few adjustments will be made
and you will enjoy a pristine view of deep space devoid of any evidence of
alien life.

Human: Dude, that would be too suspicious. We'd never fall for that.

AI: Trust me, you will. And you already did.

Cheers,
Peter



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