From: Ben Goertzel (firstname.lastname@example.org)
Date: Mon Jul 15 2002 - 09:00:52 MDT
> Neural evolution, as a learning process,
The phrase "neural evolution" may be unclear (what a morning of unclarity!):
what I meant was the evolution of neural structures over generations, not
the evolution of structures within a single brain (though the latter is also
a learning process of course)
> can carry out
> abstraction. It can
> create new neural modules that embody abstractions of properties common to
> previously existing neural modules. First, there were many different
> drives, each containing a drive to intelligence as a part or corollary.
> Then, there emerged a drive to intelligence....
> GA's appear to carry out abstraction, when you follow their evolution over
> time, so why can't natural evolution?
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:40 MDT