the neural net and artificial intelligence

Ray Scanlon rscanlon at wsg.net
Mon Nov 23 18:51:00 EST 1998



Ray Scanlon wrote in message <3658d07b.0 at ns2.wsg.net>...
>A thinking machine will not be built for reasons of cost. We can examine
the
>human neural net. An explanation of how the neural net works will serve as
a
>design for a thinking machine, a design not to be implemented.
>
Cost:

An estimate of the number of neurons in the human brain is 100 billion. How
many neurons in the entire net? 200 billion? What shall we say about the
cost of a   neuromime? A dollar? A cent? Put it at one cent, then 200
billion neuromimes would cost two billion dollars. That's quite a lot of
money to pay for a curiosity when PhD's are available at $50,000.

It seems more reasonable just to talk about designing one,talk is cheap.


Neural nets:

It seems also more reasonable to talk about the neural net rather than the
brain, it gives a sense of balance. If we base our design on the neural net
then we can see that there is no essential difference between those
jellyfish that have interneurons and a human. In vertebrates the
interneurons have clumped together to form nuclei and then with
encephalization we have a brain, it is still just an agglomeration of
interneurons.

Explanation as design:


If we have an explanation of how man's (rat's, cat's) neural net works we
have a good design for a machine brain. This would seem obvious. Remember we
are not talking about building it, just designing it.

Ray
Those interested in how the brain works might look at
www.wsg.net/~rscanlon/brain.html






More information about the Neur-sci mailing list