[extropy-chat] Be[ing] or Not Be[ing]

Hal Finney hal at finney.org
Sun Apr 18 17:38:13 UTC 2004

Harvey writes:
> On Saturday, April 17, 2004, at 08:48 pm, Hal Finney wrote:
> > "ABSTRACT. This paper argues that at least one of the following
> > propositions is true: (1) the human species is very likely to go extinct
> > before reaching a 'posthuman' stage; (2) any posthuman civilization is
> > extremely unlikely to run a significant number of simulations of their
> > evolutionary history (or variations thereof); (3) we are almost certainly
> > living in a computer simulation."
> > It is entirely FALSE to claim that the simulation argument says that we
> > live in a simulation.  It says nothing of the sort.
> I disagree.  It definitely *does* say something of the sort.  It claims 
> that if you don't agree with premise one or two, then you must admit 
> that three, "we are almost certainly living in a computer simulation."  
> It then goes on the argue that one and two are likely false, and three 
> is likely true.

You can re-write it as more of a classical argument, with premises and

1. The human species will not go extinct before reaching a 'posthuman'
2. Any posthuman civilization will run a significant number of simulations
   of their evolutionary history (or variations thereof).
3. We are living in a computer simulation.

Now you can say, the conclusion of the argument is that we live in a
computer simulation.  But you need to distinguish this from a claim that
the argument "says" that we live in such a simulation.  What it actually
says is that IF THE PREMISES ARE TRUE, then we live in a simulation.
That's different from an absolute conclusion about the matter.  It is
conditioned on the premises being true.

> > Logic and science are different, but they are both important tools for
> > reaching the truth.
> I disagree.  Unfalsifiable, unscientific methods are *not* important 
> methods for reaching truth.  Until this argument can be falsified or 
> tested or even approached scientifically, it is about as useful as a 
> religious debate about how many angels can dance on the head of a pin.

In general, logic is a way of combining information we know (or
hypothesize) to reach conclusions implied by that information.
This process is not falsifiable, nor is falsifiability relevant.
We don't believe in Fermat's Last Theorem because it is falsifiable;
it's not, because it makes an infinitely strong statement.  We believe
in it because of logical conclusions from premises we accept.

http://logic.philosophy.ox.ac.uk/main.htm is a good introduction to logic.
Tutorial 1, with its sections on arguments and validity, explains what
it means for an argument to be valid.

> I question the whole statistical assumption that we are equally likely 
> to have been born any universe or simulation, so that if there are more 
> simulations than universes we are statistically likely to appear in 
> them.  Where did these estimates of how likely it is to appear out of 
> nothing come from?  If you want to argue from unfounded intuition, it 
> seems to me that we would more likely arrive in bigger universes than 
> smaller simulations, longer-lived universes than shorter-lived 
> simulations, outer containing universes instead of inner subsets 
> configured as simulations, and in big-bang universes that actually 
> created new things, rather than in simulations which are just parts of 
> the over-universe shuffled around a bit.  Taking all the dimensions of 
> size, time, complexity, speed, and resource avaialability, it seems 
> that real universes far outweigh simulations in any meaningful 
> comparison for statistical purposes.

Now I think you are questioning the reasoning behind the argument, right?
You are suggesting that even if the premises are true, the conclusion
doesn't follow.  That's a perfectly legitimate approach, and I'm not
writing to defend Nick's argument.  I'm not familiar enough with the
details to tackle that here and now.

Faced with an argument, you can accept its conclusion; or question
its premises; or question the reasoning that links the premises to
the conclusion.  Any of these are legitimate responses.

But denying the utility of logic is going too far.  Believing that
"unfalsifiable, unscientific methods are *not* important methods for
reaching truth" will require you to reject too many useful truths.
Critique the simulation argument all you want, but do so on a valid basis.
Don't reject it just because it's built out of logic and not science.


More information about the extropy-chat mailing list