COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	Professor Philip Leith
C00006 ENDMK
C⊗;
Professor Philip Leith
Faculty of Law
The Queen's University of Belfast
Belfast  BT7 1NN Northern Ireland

Dear Professor Leith:

	Your challenge problem looks appropriate to me---rather
difficult but appropriate.  I don't imagine it will be solved
very quickly, but the Fredkin prize for winning a chess game
against a grandmaster wasn't won quickly either, and the prize
for beating a world champion is still to come.  I will write to
you more fully, when I have consulted with people more concerned
with AI and law.

	I will be happy to learn more about Ramus and the concept
of method and look forward to the book and your chapter in it.
However, precursors for ideas that didn't actually take part
in the chain of developments seem to me of limited interest.

	It seems to me that my complaints of ignorance are documented
in the article.  I complain about wrong histories of events including
mixing up (by two authors) the Dartmouth Summer Project with the
 {\it Automata Studies} volume and unjustified assumptions about
the ideas motivating the AI pioneers.  Even a sociologist has
to make an effort to find out what ideas were actually motivating,
and I don't see how he can do that without either reading papers
or interviewing participants.  Reading popular accounts by
journalists is only a way of learning how the popular intellectual
culture shared by journalists interpreted what was going on.

	I don't know what you mean by ``accept views that differ
from their own''.  If you mean ``agree with'', that's almost a
tautology.  If you mean ``try to prevent  publication'', that's
not true.  No-one in AI has threatened to boycott publishers
who also publish the works of opponents of AI.  As to accepting
papers for conferences, that's a notoriously uncertain business
when the ratio of papers accepted is small.

	As for opposing the assumptions and beliefs of AI
researchers, that's ok, but they have to be identified
accurately before they can meaningfully be opposed.

	I think Weizenbaum's quote was accurate, but I'm pretty
sure that I qualified the statement to refer to what we could
eventually tell a computer.

	Do you have any time constraints on when the challenge
problem would be appropriately published?  It occurs to me that
the next AI and law conference would be appropriate, but I don't
know when that is.

Sincerely,