COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	.require "let.pub[let,jmc]" source
C00007 ENDMK
C⊗;
.require "let.pub[let,jmc]" source;
∂AIL Professor Wilbur B. Davenport↓Department of Electrical Engineering
↓Room 38-401↓M.I.T.↓Cambridge, Massachusetts 02139∞
Dear Professor Davenport:

	While I recognize the importance of comparative scientific
evaluations in making tenure decisions, I find it very difficult
and time-consuming to do in a conscientious way.  This is why this
letter is so late.

	Pat Winston is recognized as one of the leaders of AI.  His
work as Program Chairman of the 4th International Joint Conference
on Artificial Intelligence showed a complete mastery of all the
subfields.  As Director of the M.I.T. AI Lab during a difficult
period in which the government showed an increasing disinterest
in supporting basic research, he put enormous and successful effort
into convincing ARPA to continue support of AI.  His plan for the
field was important in keeping support, not only at M.I.T. but
countrywide.

	I have been pursuing quite different lines of research
than have been followed at M.I.T. and have fallen behind in my
study of that approach.  Therefore, I was very happy to receive
a draft of his text "Artificial Intelligence".  I found it written
down more than suits my taste or consider desirable, but when I
really needed to understand the M.I.T. approach to problems of
representation and search, the information was there in a more
usable form than in any other publication.  I consider the organization
and reworking of this material a major achievement.

	Here is a seriatim discussion of the points you ask about:

	1. His engineering and scientific contributions.  He hasn't
worked in the area of formalizaton where almost all of my reading
in AI has been recently concentrated.  His thesis has a high
reputation but for me this is hearsay.  My own research strategy
has involved postponing consideration of analogy in AI until
some other matters are in better shape.

	2. Scientists of similar age
and postition in the field are Jerome Feldman at Rochester
and Raj Reddy at CMU.  I would put him definitely ahead of Feldman
and on a par with Reddy.  However, I think
Winston's scientific interests are more central to AI than Reddy's.
Winograd's thesis was more influential than Winston's, but Winograd
has also not yet been able to top it.  I think Winston has more
potential than Bobrow and Raphael who are considerably older.

	3. I don't know which M.I.T. students are his except for
Dave Waltz who did an excellent thesis.

	4. Winston shows no sign of slowing down in spite of his
heavy responsibilities as Director of the M.I.T. AI Lab.

	5. The lectures I have heard were carefully prepared and
showed a good understanding of the needs of the audience.  Therefore,
I would judge him to be an excellent teacher.

	6. His book is well written and will have great influence
on the field.

	My opinion is that granting tenure to Winston will improve
M.I.T.'s position in computer science in general and AI in
particular.

	I hope this letter is helpful.

.sgn