Re: How To Live In A Simulation

From: scerir (scerir@libero.it)
Date: Thu Mar 15 2001 - 15:24:31 MST


> Simo Kilponen wrote:
> If we were to live in a world simulated in a sense of a computer
> simulation, the detail of the sim would need to be limited and
> we would notice it somehow then. Perhaps a new way to explain
> uncertainity principle ? ;-)

In this context the meaning of randomness might be that there
is not a finite algorithm capable of computing the collapse of a
wave-function.
But rules that hold for observers who are part of a given simulation
might not hold outside of it, from a logical point of view......

> Anders Sandberg wrote:
> If I would simulate history to find out what could have happened, I
> would not just run a single simulation <.................>. So I
> would run as many versions as possible with slightly different initial
> conditions so I could get good statistics. <...............>

Computing all evolutions of all simulations is simpler than computing
just one particular simulation. That's because an algorithm running
all those simulations is much shorter. (See also the MWI of Q.Mech).
An algorithm computing just one simulation requires a lot of
informations, initial conditions, etc. It's interesting to point out
that if there are many simulations (universes) and each of these
simulations is run by a different algorithm, some selection might
occur. The best simulation (universe) might win. Which is, imo,
the simulation (universe) run by the shortest algorithm.

[See also Jurgen Schmidhuber, Algorithmic Theories of
Everything, at quant-ph/0011122, 20 dec 2000]



This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:59:40 MDT