I seem to be reading a different sheet of music from everyone else who's
taken part in this discussion. At the risk of being ridiculed for my
divergent thinking, I'd like to ask for comments, or suggestions for
reading, or answers to my questions.
For starters, the question "Why would AI want to be friendly?" seems
wrong to me. Instead, I would have asked something more along the lines
of, "Why would AI NOT develop according to the same principles as
biological life?"
As I said in an earlier post, I agree with Eugene's prediction that to
the AI humans would be part of the landscape--no more, no less. Perhaps
I was misunderstanding what he meant by "part of the landscape." To me,
it meant that humans would be simply another element of nature to the AI,
to be used or not used as opportunities come and go. (I'm defining
nature in the following way, from Webster's 9th Collegiate Dictionary--a
creative and controlling force in the universe...the external world in
its entirety) Thus, the AI would be neither friendly nor unfriendly
towards humans.
Likewise, I think I can safely say that Escherichia coli is neither
friendly nor unfriendly to humans. It can confer benefits on a human
when it lives in his gut and aids digestions of certain foods; or it can
make the human sick if he eats meat contaminated with it.
Eugene has written:
" Postbiological, as not using self-assembled heavily solvated
biopolymers and clusters thereof for self-replication and information
processing.
There's a reason why artificial life is called artificial. Not
biological. Using the same principles, but a different substrate to
apply these principles on."
The key phrase to me is "Using the same principles."
The fact that biological entities exist as nodes in webs of
interdependency doesn't seem to be particularly controversial. If you do
a www search on the evolution of cooperation you find numerous books and
papers describing attempts to model possible pathways to the cooperation
that exists in and among species--but none (at least I didn't find any)
questioning the fact that such cooperation exists.
More interesting to me is the concept of sybiosis as a basic principle of
evolution. I have often marveled at the idea that my own body is made up
of a number of different species living together, from the bacteria in my
gut to the tiny mites on my eyelashes (and the even tinier parasites
which live on the mites). At the cellular level there's further evidence
of such team efforts in the DNA of the mitochondria (and the chloroplasts
in plant cells).
If symbiosis is a basic principle of evolution (I admit that this idea is
somewhat controversial, though it seems to be gaining more general
acceptance as time goes by) and if the ability to think is a product of
evolution and if the evolution of AI uses the same principles as
biological evolution, then one would not expect the AI to be inclined to
kill off all humans or all of any other species. (It's also true that
one would not expect the AI to cater to human whims) I don't suppose
symbiosis would be an appropriate name for a pinciple governing the
evolution of a non-biological life form. Maybe co-evolution?
The recent thread on the Asian eel may be relevant to this discussion.
CYMM mentioned that equilibrium could be expected to occur at some point
in the not too distant future, and I agreed with this, writing of the
imported fire ant in central Texas. Here's a specific example of a
symbiotic type of relationship (although it doesn't fit the strict
definition of symbiosis-maybe there's some other word that would be more
appropriate here--neither cooperation nor symbiosis seem to be an exact
fit). When the imported fire ant first arrived, the armadillo population
plummeted. Armadillos nest in shallow holes in the ground, and the fire
ants were killing the young armadillos before they had a chance to grow
the tough skins and outer shell they'd have as adults. A couple of years
ago I began to see armadillos again--I don't know how yet, but somehow
some armadillos have managed to raise their young. Now a significant
portion of these armadillo's food seems to consist of the larvae of the
imported fire ant--a food that didn't exist prior to its recent
introduction to this area.
If the evolution of AI follows the same principles as biological
evolution, then I would expect AI to become, for humans, part of the
landscape--that is to say, part of nature in general. Another set of
forces to contend with, some beneficial some detrimental. I would expect
the human species to evolve in response to the presence of the AI (both
socially and biologically) but not to be killed off. What this means is
that SOME humans would die. The transition could be very uncomfortable
for the survivors. But as with all crises, there would exist opportunity
as well as tragedy.
Barbara
________________________________________________________________
YOU'RE PAYING TOO MUCH FOR THE INTERNET!
Juno now offers FREE Internet Access!
Try it today - there's no risk! For your FREE software, visit:
http://dl.www.juno.com/get/tagj.
This archive was generated by hypermail 2b29 : Mon Oct 02 2000 - 17:39:25 MDT