COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	Here are some comments and grumbles:
C00005 ENDMK
C⊗;
Here are some comments and grumbles:

Page

2.	The definition of the field of AI should not mix up applications
with the scientific goals.  I think a more precise definition of
"cognology" emphasizing its separation from psychology of humans
and animals and from electronics or computer science in general is
needed.

2.	The introduction is a bit too written down for my taste.  A
style less like teacherese will make a better impression on science
students.

2.	"Thankfully" should be "fortunately".

3.	It makes me nervous to give applications as the justification
fro scientific research, because the potential scientist may not
be interested in the applications and the bureaucrat may not see
AI as the most cost-effective approach now.  I would emphasize the
intriguing scientific problems.

	4. I don't like metaphors.  Ideas with examples is better.

4.	You use "in general" when you mean "usually".  As you know,
applied complexity theory has succeeded in finding lower bounds in
some interesting cases and this success merits a sentence.

4.	The progressive tense is "mimicking" according to Webster.

230.	Not even Feldman would call SAIL a language for AI now.
It is just a dialect of Algol with Leap data structures.
I agree about POP-2.  Again it seems to me that the salesmanship
is unwarranted.  Just tell 'em.

248.	Using SETQ with more than two arguments works in MACLISP,
not in Stanford LISP.

257.	I don't find the use of apply in redefining countatoms elegant.

259.	You should also refer to Stanford LISP and its modification
UCI Lisp and at least one LISP for the IBM 370.