Re: Infinitely fast computer

William Kitchen (bill@iglobal.net)
Fri, 11 Oct 1996 21:59:53 -0500


Hal Finney wrote:
>
> An idea I've been amusing myself with a bit, relating to the question
> of how hard it will be to generate AI using nanotech is this: suppose a
> genii gives you an infinitely fast computer. .......
>
> The question is, given such a miraculous device, how hard would it be
> for you, meaning the typical programmer reading this, to produce a program
> which could pass the Turing test, or better still one which is super-
> intelligent? Where would you start? How long would it take you to write
> the code? What research would you have to do? Could it even be done?

This is indeed an interesting thought experiment. My suspicion is that
the
time between being granted such miraculous machinery, and the emergence
of superintelligence would be exceedingly short. On the order of weeks
for one such as myself who only programs occasionally, perhaps only days
for an experienced programmer who does it as his primary method of
making a living, possibly even hours for a first class programming
wizard (especially one with a passion for mathematics).

The reason I think that such a short time scale is possible is that the
feat could perhaps be achieved by creating a program, or more likely an
interacting group of competing programs, that evolves on it's own once
set in operation. With unlimited computing resources (or at least
enough to seem unlimited by modern standards), this initial population
of virtual life would not need to be particularly efficient or effective
in it's evolutionary progress. It would only have to meet these
requirements:

1. it could not permanently stagnate at any particular level of
development
2. it could not get into any kind of degenerative "devolution" mode
(or at least, such occurrences would have to be outweighed by
development in a positive direction)

With truly immense computing power to play with, reiterative testing and
modification of starting populations would be easily accomplished, so
all false starts could provide valuable clues towards a functional
solution. And the first initial population to successfully achieve the
above conditions would be an instant success. With infinite computing
resources, a superintelligence would exist as soon as the first
successful program is set in motion. It seems to me that the only
limitation of the development of such an intelligent virtual lifeform
would be imposed by the limitations of the available computer. I cannot
concieve of what kind of intelligence could evolve in a computer that
was infinitely fast and with infinite memory. But even with very large,
but still finite resources, it is easy to imagine that the potential for
such a virtual life scenario is awesome.

-- 
                              Peace,
                              William Kitchen

bill@iglobal.net