COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\\M0BDR25,\M1BDI25,\M2BASL30\.
C00006 ENDMK
C⊗;
\\M0BDR25,\M1BDI25,\M2BASL30;\.
\F2\CSTANFORD ARTIFICIAL INTELLIGENCE LABORATORY
\CDEPARTMENT OF COMPUTER SCIENCE
\CSTANFORD UNIVERSITY
\CSTANFORD, CALIFORNIA 94305
\F0





							March 4, 1974



Editorial Staff
\F1San Francisco Chronicle\F0
Fifth and Mission 
San Francisco, California

Editor -

\J	Professor Schwartz's March 1 letter stating  that I was naive
in believing that  the work of the Artificial Intelligence Laboratory
"does not have immediately discernible military applications" reminds
me that I  don't agree either with the \F1Chronicle\F0  article (Feb. 4) or
with Professor Schwartz.

	In  my opinion, work in artificial intelligence has potential
military application  and  the example  Professor  Schwartz gives  of
improving remotely piloted aircraft will  do as well as any.  It also
has important civilian applications \F1e.g.\F0 making an automatically
driven  car that  can be  used  safely by  children  and old  people.
Somehow,  the \F1Chronicle\F0 reporter  got the mistaken  impression that I
thought the  Defense  Department does't  know  what  it is  doing  in
supporting research in artificial  intelligence.  They are definitely
aware of  the possibilities, although they are also aware of obstacles
to immediate application.

	My quarrel with  Professor Schwartz is  probably deeper.   In
the  first place, our  work is  motivated primarily by  the scientific
interest of the problem of intelligence and by civilian applications.
Most of  us don't feel  that the  country is in  enough danger to  be
motivated primarily  by military applications.  However,   I do think
that a development that strengthens the United States relative to the
Communist countries is  a good thing rather than a  bad one, but this
is  a political rather than a technical  judgment.  Secondly, I think
that automatically piloted aircraft can be made to distinguish better
between civilian and military targets than can humans under stress.

	Finally,   I think Professor Schwartz is  myth-making when he
imagines that "quantitative  decision theory"  is going  to take  the
decisions about whether to go to war or not away from our political leaders.
Our political leaders are  not about to give up any of their power to
any automatic process.\.



							Sincerely,


							John McCarthy
							Director, Artificial Intelligence Laboratory