COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	%giving[f88,jmc]		Giving Mental Qualities to Machines
C00005 ENDMK
C⊗;
%giving[f88,jmc]		Giving Mental Qualities to Machines

	Philosophers study question about mental qualities: What
are beliefs and other mental qualities?  Can machines have mental
qualities?

	From the point of view of artificial intelligence, the
questions are somewhat different: What beliefs should we give our
computer programs?  What should the beliefs be about?
How should they be represented in computer
memory?  How should they change as the program operates?
What beliefs and mental qualities should they ascribe to people
and to other computer programs?

	The philosophers don't think about mental qualities this
way, so it's hard to get much out of their 2,000 years of labor
that is helpful for AI.  Hard but not impossible.

	For example, consider the idea of {\it natural kinds}
recently treated (and maybe originated) by Hilary Putnam.

What clarification can we get from philosophers that is useful
in AI?

	1. anti-behaviorism

	2. natural kinds

	3.

Treat the central cases first.  Thus we must first understand
``Tom wants to have an affair with Bill's wife'' in the case
when Tom admires a particular lady and correctly believes she
is Bill's wife.