The universe as computer

[Russell Turpin]

Mr. Fork:
>By 'perfect' do you mean one that is not different?

That's a good question. In part, the exercise might be
to define what quality a simulation must have to be
considered (in what ways?) equivalent to the "real"
thing. Putting on my hat as an expert in simulation
modeling, I would say that a simulation gives perfect
answers when (a) every operational question about the
modeled system is answered by the simulation, and
(b) there is an operational way to distinguish the
aspects of the model that reflect the modeled system
from those that are simulation artefacts, i.e., that
are irrelevant to the modeled system.

This notion of perfect excludes performance. We do
simulations because they are cheaper or faster or
more practical than examining the modeled system. In
the real world, no one builds or wants a simulation
that is perfect in the sense just described. Instead,
they want an answer next week that is "good enough,"
without having to first spend a billion dollars on
actual experiments. For this philosophical puzzle, it
seems to me the first requirement is perfect answers.
But feel free to define "perfect simulation" in some
other way. The real question is: When do you consider
a simulation equivalent to the real thing? And then
there is the related question: How do you know we're
not in a simulation, and when would it matter?




_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com