From: Clark Graham (ggc198@ecs.soton.ac.uk)
Date: Wed Mar 07 2001 - 19:55:06 GMT
http://cogprints.soton.ac.uk/documents/disk0/00/00/03/19/index.html
> Hudson:
> I think we would still have a problem building one in any way other
> than to create it in its infantile state and to let it grow and
> develope as we do. How can you have a mind that is recognisable as
> human without a personality and a large collection of memorys? And
> how can you get those without experiencing life as we know it?
Clark:
Building an infant / baby rather than a fully grown adult does seem to
be the logical thing to do (if only just to start with). Aiming for a
complete adult first time seems to be a bit like trying to build
concorde before a simple bi-plane. But suppose we do eventually
understand the human mind, including how memories are stored. Maybe
then it would be feasible to program a "base set" of memories into an
artificial "being", as well as an initial set of synaptic weights for
its "brain", and in doing so starting them at a fully-grown age rather
than as a new-born child.
> Hudson:
> Also how could we possibly verify that 'someone was home' in our
> creation? We can't even do that with each other.
Clark:
If we can't verify this with each other, then why should we expect to
be able to verify this with a machine? If it could act, feel, etc just
like us, then do we really have the right to say that "no-one is
home" (or vice versa)? Although we can't test for consciousness
explicitly, we can test for some of the results of it, ie. "human"
characteristics. This is done via the (robotic, probably) Turing Test.
Graham.
This archive was generated by hypermail 2.1.4 : Tue Sep 24 2002 - 18:37:21 BST