X-Message-Number: 11654
From: Thomas Donaldson <>
Subject: First, a general reply
Date: Sat, 1 May 1999 23:28:24 +1000 (EST)

Hi everyone!

The point I am making may be subtle; so that I shall explain it once more.
Since there have been several replies, I will try to deal with each one,
but that might easily get difficult just because the replies are different
and may each require their own long explanation.

But regardless, here goes:

First of all, it seems to be assumed by several of those replying that
an emulation of the world by computer will someday become possible ---
at least, an emulation good enough that we are fooled by it, not in a
temporary way (after all, there are programs with make conversation and
do fool people right now). A major point I was making was that such 
emulations entirely within a computer are highly unlikely to be as good
as the real world itself. The major feature which the real world shows
is that over time it becomes effectively infinite --- and when examined
at high resolution, the same happens. Computers are finite; but I have
no doubt that at any time in the future there will continue to be truths
about the universe which we do not know. And when we notice such truths
(whether they are otherwise trivial or not) they will be facts that 
the emulation did not include. 

The work on joining general relativity with quantum mechanics is even
part of this (in its own glorified way): the current ideas are not that
the world is ultimately discrete, but that it becomes ultimately fuzzy.
These are not the same things at all. (Though current ideas may be
shown to be inadequate in any case).

Sure, we might have an emulation which works for a short time. The real
problem is that of making one which works into the indefinite future. The
differential equations don't even converge to a single solution: they
are nonlinear and so produce a variety of possibilities depending on 
just how accurate the initial values may be. And even in discrete form
the same will happen... if anything, to a much greater degree.

The next issue comes when we consider just what kind of device might
work with the real world rather than a computer emulation. The very first
point is that such a device simply cannot be sequential. We are not
pure brains isolated completely from the universe. We must constantly
respond to it. Even if we suppose (as I understand Mike Perry to be 
supposing) that there is some system which takes input from the real
world and conveys it to our slow sequential Turing machine, that 
Turing machine remains totally unable to respond to any changes in the
world fast enough to last more than a few hours. Can it cross a road
with automobiles moving each way on the road? Can it even get on a bus?
Too many things are happening around us, all at once, for us to be 
emulated by a sequential computer.

There is also a second problem, either more profound or more simple. 
I will call it the Black Swan problem for short. We work with our
language, and therefore with our thinking, in an open-ended way which
DEPENDS on input from the world even for us to get an idea of what
a swan IS. We do not think within a closed system such as a set of
axioms and their consequences, but change our definitions and ideas
about reality in response to what we perceive. And without that input
from the world, we basically become inactive, or fall asleep, or cease
to respond in other ways. Moreover, it is the UNPREDICTABILITY of that
input which keeps us active. When people saw black swans, they did not
invent a totally new word for these creatures; they decided that they
were seeing black swans --- despite the definition of swans in every
dictionary as white.

Please note that (as I've said before) I am not discussing here the
impossibility of creating a nonbiological but aware creature. I am 
discussing what it means to be aware. And awareness, in the way in which
we are aware, requires that we be aware of something which cannot be
emulated simply by a computer. That something will always have an element
of unpredictability to it, good or bad --- and that unpredictability 
comes from a form of ORDER, not simply from randomness.

In my next messages, tomorrow, I will deal with the various objections
individually.

			Best and long long life,

				Thomas Donaldson

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=11654