Re: Simulation argument in the NY Times

From: Peter Butler (
Date: Fri Aug 17 2007 - 07:24:30 MDT

> Actually... considering your original question with the more liberal
> definition of simulation, perhaps it is not so unlikely as I claimed,
> that say a turing machine enumeration might appear spontaneously.

It may be no more implausible than, say, the spontaneous organisation of
proteins into self-replicating structures. As you say, assigning
probabilities to these is difficult, to say the least, especially
considering the fact that we have no knowledge of the conditions that
occur in the universe which contains ours. Can we be sure that the logic
required for the execution of a Turing machine exists in the containing

I think that this question needs to be settled empirically. We need
some way of determining whether in fact our universe is a construct
running on some other substrate. Once this is determined to be the case
we need some way of deriving information from the containing universe.
However, this may be problematic, in the same way that figuring out from
within an operating system that it is running on emulated hardware is
problematic. Problematic, but not impossible.

> Of course, conway's game of life is ITSELF a deliberate simulation.
> The likelihood of a deterministic cellular automata arising naturally,
> compared with our kind of universe, is unknown, since we've never seen
> a deterministic system arise naturally, and under the simulation
> hypothesis, we've never seen our type of universe arise naturally either.


> How can we assign probabilities to extrapolations when we have no
> idea what proportion of the complete set of imaginable things we
> can imagine?
> This objection applies to every assignment of probability, not just
> ones related to looking outside the box. I'm not sure what the real
> probability answer is, if there is one.

Not to every assignment of probability, just to the ones where we have
incomplete information - for example, I know that the dice is fair, I
know Newtonian physics, so I can be reasonably certain that the chances
of rolling a particular number are about 1 in 6.

To me, the incompleteness of the information available about our own
universe is the crux of the problem: we simply don't know enough to say
for certain whether we are living in a simulation, much less to make any
guesses about the nature of the universe which may or may not be
simulating us. Needless to say it would be very interesting to find out.


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