Re: the Turing test

Dr. Rich Artym (rartym@galacta.demon.co.uk)
Sat, 26 Oct 1996 15:50:33 +0100


I was going to write:

> No! You're *all* AIs, and I'm the only live human here!

And then suddenly I realized the truth. We're all AIs ... :-)

'Come to think of it, that's not entirely a joke. As far as we know,
nature doesn't actually create intelligence, but only an infrastructure
that can develop intelligence through interacting with its environment.

In that sense, Mankind develops its own intelligence, so it's artificial
despite being implemented on a biological substrate. There's more to
this than just a play on words, because if that view is true then we are
only a few short steps away from creating an artificial substrate which
can develop intelligence in the same way as we. The neural technology
is more or less here already, especially now that multi-valued systems
are no longer frowned upon except by a few academics, so all we need is
micro-scale replication (nano not needed in principle, but possibly yes
in practice) for such substrates to become practicable.

Question: has the AI community run a learning neural automaton long
enough to simulate even a fraction of the learning process of a typical
child by the age of (say) three? I think I know the answer ("No", by
several orders of magnitude because of the computational requirements),
but confirmation or denial from within that community would be useful.
Of course, even that computational epic would probably fall short of
producing a child mentality, but it would provide useful information
towards that goal.

Rich.

-- 
###########  Dr. Rich Artym  ================  PGP public key available
# galacta #  Email   : rich@galacta.demon.co.uk         158.152.156.137
# ->demon #  Web     : http://www.galacta.demon.co.uk   194.222.245.150
# ->ampr  #  AMPR    : rich@g7exm[.uk].ampr.org 44.131.164.1 BBS:GB7MSW
# ->NTS   #  Fun     : Unix, X, TCP/IP, kernel, O-O, C++, SoftEng, Nano
###########  More fun: Regional IP Coordinator Hertfordshire + N.London