COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	review[w88,jmc]		Review of The Question of Artificial Intelligence by Bloomfield
C00006 ENDMK
C⊗;
review[w88,jmc]		Review of The Question of Artificial Intelligence by Bloomfield

Notes:

Does Shanker assert the impossibility of a a specific performance?
Is there some biological or psychological understanding that he
denies?

It is a pity that Shanker ignores the areas in which AI research
overlaps philosophy and in which AI research has profited from
the work of philosophers.  Consider the question of how knowledge
differs from true belief.
Gettier examples.
Frege
Positivism as a doctrine concerning how to construct intelligent
programs.

Philosophers at the crossroads.  Will they use their 2,000 years of study
of important questions or will they stand on the sidelines and
see epistemology detached from philosophy?

Some will do one thing.  Others will do another.

We don't want to eliminate philosophy from the science of AI ---
at least not so long as there is some hope of getting philosophers
to do some of the work.

knowing that, knowing how, knowing what, knowing whether,
knowing about, all I know is

19 - as measured by the spreading influence of the paradigm.  No, by
the fact that a modern chess program can beat S. G. Shanker.

21 - ``It shows that now factual claims have been made.''


queries to look up
Does Kowalski refer to McCarty, so that Leith ought to have noticed?
The baleful influence of Tony Battista.

Even Shannon didn't refer to information theory in his one article
on AI --- the paper on chess.

ALLEN.NEWELL@A.CS.CMU.EDU
history request
In a review of a bad book, The Question of Artificial Intelligence,
I plan to make some remarks about the institutional history of AI including
something like the following.

"AI didn't start as an establishment.  Minsky and I were fortunate that the MIT
Research Laboratory of Electronics had a ``joint services contract'' that
permitted its head, Jerome Wiesner, to say yes instantly when we
encountered him in the hall in May 1958 and asked for a secretary, a key
punch, and two programmers.  He countered by asking if we wanted to
supervise six graduate students in mathematics whom he had committed
himself to support but had no immediate job for.  I believe that the
Newell and Simon work at Rand started in a similar informal way."

Is my conjecture correct or was there a formal proposal to begin your
and Herb's work on complex information processing?

Shannon, C. E., "Programming a digital computer for playing chess,"
Phil. Mag., 1950, 41, 356-375.