"J. R. Molloy" wrote:
> Franklin Wayne Poley asks,
> > What do we mean by "brainpower"?
>
> Primarily, I think the term refers to the ability to solve problems or answer
> questions.
>
> --J. R.
It is very difficult to pass these concepts out to the public while maintaining
the vital context. To have the power to solve complex problems we need a
combination of processing units, and an effective arrangement of such units.
Back in the 50's people observed that it was unreasonable to expect to build a
computer out of millions of vacuum tubes (the basic units of processing) to rival
the human brain, even if only a tiny fraction of the neurons of the brain were
needed to do thinking. This is a strong argument, even though it does not really
prove anything. It may be possible to reproduce the problem solving power of the
brain with just the right arrangement of vastly fewer processing units.
As time has gone on, the number of artificial processing units has grown
exponentially. At the same time, biologists have found vastly more cellular
processing activity. The point of all this is not to say that when the
artificial processing capability exceeds the cellular processing capability we
will have intelligent machines, but rather, that at that point the "we will never
have an equivalent number of vacuum tubes" argument goes away. This is the
removal of a negative, not necessarily the demonstration of a positive.
Although we want to jump to the ultimate conclusion, what Moravec is actually
saying is that the extrapolations of the trends show that soon you will have one
less reason for calling his work "unreasonable." It does not say anything about
when we will, or will not, have intelligent machines.
-Ken
This archive was generated by hypermail 2b29 : Mon Oct 02 2000 - 17:36:27 MDT