What's Up

     I just had a neat though, and I wanted to get it down on
paper. One of the highest aspirations of the computer
programming community is to create so-called "artificial
intelligence". This would be, of course, a computer that can
think on its own, make decisions, have opinions and
intuition, and be more than just a fancy high priced
calculator. In short, it would emulate the behavior of a real
human being. This idea has also been the focal point of
many science fiction stories; where some computer
becomes "self aware" and tries to obliterate all of humanity.

     While all of this is going on in the background, at the
same time, there are literally hundreds of millions of real
humans interacting in real time on the Internet. What if a
computer program could model synthetic human interaction
based on the activities of all of these hundreds of millions of
real humans all using the Internet at the same time?

     I'm sure you have heard the term "collective
conscienceness". That's the idea that everyone's personality
and interaction adds up to one John Q. Public. Our
capitalistic society has relied on this concept for over a
century. With a model like this, some lady in Boise, Idaho
picking out a drapery color for her breakfast nook might play
a role in making a decision that would effect government
policy! Sick thought? I don't know. One thing that makes this
idea different is that the intelligence of the system would be
totally dependent on huge numbers of living humans. That
means that the computer would never be independent of
human life and never come to realize that we humans are
quite useless. Also, the computer's collective intelligence
would evolve and develop with that of all of humanity. [Insert
your own joke here.] This might sound like an impossible
task, but then consider the kind of computing power that it
takes just to catalog all of the millions and millions of pages
on the World Wide Web and present them to you in a
searchable database.

     Currently complex systems of pseudo artificial
intelligence are being used to forecast everything from the
weather to the stock market. There are also programs that
learn and evolve over time. OCR or optical character
recognition is used to translate a picture of written words on
paper to text that the computer can store as though
someone had typed it in. The computer can read! The more it
reads and gets corrected for its mistakes, the better it gets.
Recognition of human speech is also being developed. Even
though a computer's ability to read and translate spoken
language into text might require techniques of artificial
intelligence, this still does not make the computer sentient. It
can't really do anything of its own will with this information.

     Perhaps from all of the billions of instances of real
human interaction on the Internet a computer could someday
decant just exactly what the average person with real
emotions would do. With the accumulation of thousands of
years of written history and learning AND the ability to make
human like decisions, the computer could develop a totally
non biased intelligence that is far greater than that of any one
living human. Then what would we do? JL :o)