INFORMATION TECHNOLOGY SYMPOSIUM
211
further diversity.
If
everything becomes ordered and there's no chaotic
environment in which the evolutionary process takes place, then you
would saturate the environment and that process would break down.
The Second Law of Thermodynamics states that disorder, called
entropy, randomness, or chaos, is constantly increasing. And in fact,
evolution is a very paradoxical result: you have this increasing order and
this very elaborate, intricate structure that emerges in an environment
which the Second Law of Thermodynamics says should become increas–
ingly disordered.
Well, all of life, including all of our technology, is insignificant in
terms of the amount of entropy and chaos in the environment; and the
Second Law of Thermodynamics is providing ever greater chaos in the
world to provide the options for further diversity. So we're certainly not
in any danger over the next century, or couple of centuries, of running
out of the increasing chaos in the environment, and therefore this
process can continue. The order of the technology itself is yet another
resource; as long as you have evolution, you have the order that it pro–
vides for the next generation. Now conceivably, there is a point at which
you saturate all matter and energy with very highly evolved intelligence
and run out of chaos in the environment. But if you examine that mathe–
matically, the amount of chaos that exists, let's say at the quantum level,
is so vast, even if at the point when you've created intelligent entities
that are trillions and trillions of times more powerful than all of human
intelligence today. Maybe it will saturate at some point, but way beyond
our ability to conceive of intelligent entities today.
Audience Member:
What is your technical background?
Ray
Kurzweil:
My technical background is something called pattern
recognition which is part of artificial intelligence. It's building systems
that recognize patterns, like printed letters or speech or objects on the
table, and so on. And the methods we use are in fact what Edward
Rothstein referred to, which are models of biological systems. They are
chaotic methods; one method is called neural nets, which is based on
these mathematically simplified models that we think of as networks of
biological neurons. Another method, called evolutionary algorithms,
emulates evolutionary processes in the computer. Sometimes called
genetic algorithms, these systems share the chaos and unpredictability of
the biological world because they are attempts to model the biological
world. But that doesn't mean the decisions they make or the equilibri–
ums they reach are meaningless-quite the contrary.
