Re: Emulation vs. Simulation

From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Tue Mar 27 2001 - 17:12:09 MST


Eugene, I think, and then Emlyn pointed out my droppings on the carpet
by attempting to compress the data storage requirements for behavioral
input-to-output mappings to something reasonable.

I'll claim in my defense that I've outlined what I believe the size
of the available lookup tables might be for independent entities running
in reality as well as their communications range for teleoperation
with much larger lookup tables. If you want to disprove my position,
you should consider a robust exploration of how many bits humans are
exposed to and how many are really required for the behaviors people
exhibit as outputs on a daily basis.

For example, if over a human lifetime, you can speak approximately
10^9 words, and there are ~10^6 possible words you can speak, that
gives you a lot of combinations, but if you compress it based
on eliminating the meaningless statements (probably a very large
part) and then use compression to reduce the storage requirements
of the most commonly spoken phrases or word combinations. Anyone
care to convince me that those storage requirements exceed the
storage capacity of a human body that I've previously outlined?

E.g., On Wed, 28 Mar 2001, Emlyn wrote:

> Too true... if you start using heuristics to chop out chunks of the table,
> you move back toward an imperative algorithm for consciousness, which
> undermines your argument. The argument concerned a pure lookup table, which
> did not have any iterative processing involved...

BeeBaDa BeBaDa BeBada phooey!

I can use fuzzy logic to compress the size of the lookup tables
and only advance a micro-step back towards "consciousness".
I've defined precisely what I consider the elements of consciousness
to be in another message in this tread. Compressing the lookup
tables and using fuzzy logic doesn't get you anywhere *near*
fullfilling those criteria.

So -- bifflesticks.

Robert



This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:59:43 MDT