COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\\M0BDR25\M1BDI25\M2NGR30\M3XMAS25\.
C00008 ENDMK
C⊗;
\\M0BDR25;\M1BDI25;\M2NGR30;\M3XMAS25;\.
\F2\CARTIFICIAL INTELLIGENCE LABORATORY
\CCOMPUTER SCIENCE DEPARTMENT
\CSTANFORD UNIVERSITY
\CSTANFORD, CALIFORNIA 94305
\F0

\C6 April 1973


Mr. Wes Gallager
President and General Manager
Associated Press
50 Rockefeller Plaza
New York, N.Y. 10020

Dear Mr. Gallagher:

\J	Here is the present situation with regard to the points mentioned
in your letter of 29 March.

	1. Our request for a grant has still not been acted on by the
National Science Foundation.  I gather that some kind of bureaucratic
tangle has prevented them from even considering whether they regard the
proposal as having any merit.

	2. If they give us the grant we request, there will be somewhere
between ten and twenty terminals in homes and possibly on the Stanford
Campus.  Some kind of news service would only be one of the services
offered.  We will not proceed further with our planning until we get
a positive response from a potential grantor.  What is the possibility
of the Associated Press sponsoring research in home news services?

	3. No-one presently uses our APE program for any purpose for
which they might be likely to pay, so that AP is not subsidizing
competition with itself or any of its clients.  I gather that you have
some worries beyond this, but I don't have a clear idea of what they
are.

	4. The present usage is about half within our laboratory and
about half from other laboratories on the ARPA network like the M.I.T.
laboratory you visited.  There are about thirty laboratories presently
attached to the network.  Much of the usage is probably simply demonstration
of the network and of the existence of such a program.  The other usage
is from people satisfying personal curiousity about the news.  There is
some desire to use the news to experiment with more sophisticated ways
of getting information from the news than a simple keyword search.  However,
such a project has not yet materialized.

	5. We expect to have more precise information about usage when
we put a questionnaire about usage into the program.
When we get it, I'll send you a copy of the report.
	
	6. We received an inquiry from the UCLA student newspaper about
whether they could get AP news through the ARPA net.  If we had wanted
to experiment with this, we would have asked your permission, and perhaps
they could have paid you.  Certainly, it would not have been in accordance
with our agreement to do it without further permission.  I don't know how
they proposed to get access to the net which also costs money, so I suspect
that it wouldn't have worked.  We didn't want to do it anyway, because our
computer is already overloaded, and it didn't seem as though there would
be a sufficient scientific payoff.

	7. It occurs to me that there might be a commercial demand for
a computer based news service that would cost a couple hundred dollars
per month and which would be more convenient than the news teletype itself.
Government agencies and businesses might want it, and news services like
AP might be interested in supplying it.  The capital cost of setting it
up would be a few hundred thousand dollars or less if a present time-sharing
computer service bureau had a suitable computer.  If you are interested,
I suppose we owe you a bit of free consulting for causing you worry.

	Again let me assure you that we shall request your permission
before changing the way the AP service is used in any way.\. 


				Sincerely yours,



				John McCarthy
				Professor of Computer Science
				Director, Artificial Intelligence Laboratory