Re: Simulation argument in the NY Times

From: Larry (
Date: Wed Aug 15 2007 - 09:40:21 MDT

On Tue, 14 Aug 2007, Eliezer S. Yudkowsky wrote:

> Robin Lee Powell wrote:
>> Like say, for example, floating point errors at the extreme margins
>> of the large and small?
>> Would look a lot like quantum uncertainty, don't you think?
> Wouldn't look anything remotely like quantum uncertainty. For one thing you
> couldn't do quantum computing with it.

The simulation would absolutely have to be on a quantum computer.

Even though we don't have any decent working quantum computers yet, we see
many things which absolutely depend on the insane computing power available
in reversible quantum systems. A hologram is a good example, fire one
photon at a time through, you will still get the image built up slowly.
That one particle has had to do a very complex quantum computation to
correctly "choose" the right destination.

To even simulate picking up a hologram and looking at it under room light
would take utterly insane amounts of classical computing power. It just
gets worse, the bigger the scale the more possible paths, the more
possible particle entanglements, etc.

A fun thought about classical vs. quantum physics. I may be wrong here,
but it seems to me that our semi-classical world with locality arises out
of entanglement. Entanglement has an 'if then' nature to it, if X is true
about A, then Y is true about B. This would seem to give rise to locality
and space "if A is around (x1, y1) then B is around (x2, y2)". Of course
unlike classical mechanics, these 'if then' statements are non exclusive,
"if A is around (x1, y1) then B is around (x2, y2) or (x3, y3)".

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