Re: Simulation argument in the NY Times

From: Vladimir Nesov (
Date: Sat Aug 18 2007 - 03:14:30 MDT

Saturday, August 18, 2007, Norman Noman wrote:

NN> It does affect the insider's experience, because not all simulations are
NN> hermetically sealed. If we are in one which is not, what the interface is
NN> and what is outside will matter, and they will affect our experience. I'm
NN> really not sure what you're trying to say.

I see. I talked only about air-tight simulations. Set of such
simulations includes simulations that from the inside are equivalent to
those which are interacted with. OK, I might have unnecessarily
obscured things trying to make informal illustrations. Say, there's a
simulation history X, which includes all subjective experience
obtainable from within a simulation up to a moment T (internal
simulation time). A and B are universe models that generate X (that
is, they are specifications or program code that when executed
produce X during first T), but they differ after that. If you are a subsystem of X and you know
that there're 1000 of implementations of A and 1 of implementations
of B currently running in the 'real world', your expectations of what
will happen after T are not affected by that informations about
number of implementations. If you know that all these simulations are
not going to be disrupted, you shouldn't care about number of
instances, since they can't interact with each other. If
there's also a single all-enumerating TM, both A and B have
an implementation.

 Vladimir Nesov                  

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