Re: Simulation argument in the NY Times

From: Norman Noman (overturnedchair@gmail.com)
Date: Wed Aug 15 2007 - 21:32:56 MDT


> Most 2D games still have the APPEARANCE of a third
> > dimension, and most 3D
> > games have action confined to a surface. It's fairly
> > clear in both cases
> > that they are made to be played on a 2D screen, and
> > thus by 3D people.
>
> Hmmm.. according to the string theorists, space-time
> is ten-dimensional. Therefore, it's meant to be played
> by eleven-dimensional creatures. Maybe there's a cheat
> code hidden somewhere in the sixth dimension?

Most of the dimensions are small enough that they are functionally
nonexistent on the macroscale. If the universe were a video game, it would
easily be playable on a 2D screen.

> Due to the nature of the
> > > singularity, their existence is currently
> > impossible to describe.
> > >
> >
> > It seems to me the very fact that they are
> > "post-singularity level beings"
> > is the beginning of a description.
>
> So is calling something a "thingy". So is calling
> something an "n-rank tensor in k-dimensional space".
> Trying to get useful information from the statement
> "post-singularity level beings" is on roughly the same
> order of absurdity as trying to derive general
> relativity from knowing "the curvature of space-time
> is an n-rank tensor" (and nothing else).

I'm just saying, the idea that things are "impossible to describe" seems a
like a bit of an exaggeration to me. It's an H. P. Lovecraft cop-out.

"What is water? It's a difficult question, because water is impossible to
describe. It's sort of like birds. What are birds? We just don't know."
- Look Around You

> while a
> > post-singularity entity could be
> > and do almost anything, what it WOULD do depends on
> > the original
> > optimization target of whoever orchestrated the
> > singularity, and whether
> > something went wrong and that target became
> > corrupted.
>
> Yes, but we don't know the specifics of how any of
> this happened. You might as well try and solve for a
> billion unknowns with one equation.

It's certainly difficult, but functionally different answers could be
enumerated, and their relative probability estimated.

For instance, it is unlikely the universe was simulated to solve the math
problem "what is six times seven" even though the solution to this problem
exists within the universe, because simulating a whole universe is not an
efficient way to do simple math.

It's POSSIBLE. But it seems to me less likely than that the universe was
simulated, for instance, just to see what happens. Wouldn't you agree?

> In either case, the original target was probably
> > something we would not find
> > totally alien, if it was the target of evolved
> > beings, because evolution is
> > likely to create beings that are like us in many
> > ways, regardless of the
> > environment. Beings that like surviving, eating, and
> > reproducing. Beings
> > that are just exactly smart enough to create
> > civilization.
>
> How do we know that these creatures *evolved*?

If they were designed, then they were designed according to the motivations
of those designers, and so on, until we get back to creatures who were not
designed. I suppose it's possible that they just spontaneously sprang into
being, or that their brains and bodies were formed by the accretion of
sediment layers, but evolution seems more likely.

How do
> we know that the Second Law of Thermodynamics operates
> in their universe? How do we know that one plus one
> would equal two?

These things are part of the complex fundamental logic of our universe, to
which no alternative is known, or is even possible for us to simulate. The
likelihood that the world outside the system operates by a system of rules
which is known to exist is greater than likelihood it operates by a
hypothetical system which is not.

Additionally, order is known to beget order. A vast surrealist nightmare is
not.

You can dream up whatever kind of beings you want, so
> long as they have adequate technology, and then cook
> up a plausible-sounding argument for why they would
> simulate us. To shamelessly steal Eliezer's
> terminology, the hypothesis of a "universe simulation"
> does not forbid any outcomes; it does not serve to let
> us anticipate the future. It is a disguised hypothesis
> of complete ignorance.

It seems to me much less likely that a pre-singularity civilization would
simulate a universe than a post-singularity one, I just think it's a
possibility which should be explored further before being discarded.

While the simulation hypothesis does not forbid any outcomes, it gives them
different probability weights. I disagree with Eliezer that it is a
disguised hypothesis of complete ignorance.



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