COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\\M0BASL30\M1BASI30\M2BASB30\M5NGR40\M3NGR25\M4NGR20\MFSTA200\
C00015 ENDMK
C⊗;
\\M0BASL30;\M1BASI30;\M2BASB30;\M5NGR40;\M3NGR25;\M4NGR20;\MFSTA200;\;
\'3;↓↓\FFS\F5
\'3;↓Q\CSTANFORD UNIVERSITY
\F3\CSTANFORD, CALIFORNIA 94305
\F4



ARTIFICIAL INTELLIGENCE LABORATORY\←L\-R\/'7;\+R\→.\→S   Telephone:
COMPUTER SCIENCE DEPARTMENT\←S\→.415-497-4430
\F0\C23 December 1974



Editor
\F1Communications of the ACM
\F0ACM Headquarters Office
1133 Avenue of the Americas
New York, New York 10036

Dear Sir:

\J	I wish to take issue with Lawrence Press's \F1Arguments for
a Moratorium on the Construction of a Community Information Utility\F0,
which appeared in the December 1974 Communications.  Before disputing
points explicitly made in the article, I would like to dispute some
apparent assumptions that are not explicitly made.

	Assumption 1. A community information utility is a natural monopoly.
Press's sketch of a CIU involving the cable TV system
suggests that CIU is a natural monopoly of the cable TV operator,
although he doesn't say it.

	This would be unfortunate if it were true,
because the information services provided would be
auxiliary to the TV business and would develop slowly, but consider
the following alternate scenario:  

	\F1Home terminal services are offered by time-sharing companies
as soon as terminal costs come down enough, useful services are developed,
and selling by advertising rather than selling by hand-holding brings
the sales costs down.
The user owns his terminal and there are many computer systems competing
for his business.  The communication is by dial-up telephone;already with
concentrators in local dialing areas, communication is cheap enough so
that national competition is possible.  The purveyors of information
services are distinct from the operators of the time-sharing systems
so that the capital cost of creating and offering for sale a new information
service is not large.  (This seems to be happening naturally, but if
obstacles develop, the Sherman Act can be applied.)  It may happen that
the creation of a CIU is not a discrete event, but more and more services
and information files become of interest to more and more people with
less and less interest in computing per se.  For example, airline and
theater ticket reservation services will increasingly make their systems
available for direct public access.\F0

	Assumption 2.  Information Services must be designed for the
whole community.  Like every innovation, home information services
will initially be expensive, awkward to use, and unreliable.  Therefore,
they will first be adopted by the well-to-do and by people who
find them attractive enough to devote more money and trouble than
the average person will.  The initial marketing problem is to attract
such a clientele.  This can happen as soon as good display terminals
for $500 are developed and competition forces the time-sharing companies
to cut excessive overhead.

	Assumption  3.   It  is  possible  and  proper  for  computer
scientists, humanists, sociologists and writers to decide what if any
community information services would benefit the public and  for  the
government to make available those found desirable and forbid others.
In my opinion, the inability of these groups to evaluate the \F1future\F0
consequences of an innovation is indicated by their inability to
evaluate comprehensively the manifold \F1present\F0 ways people use
automobiles.  It also expresses a certain contempt for one's fellow
Americans to assume that they are such bad judges of whether it is
in their own interests to use an innovation that the experts must
decide for them.

	Assumption 4.  The development of community information services
by companies trying to make a profit would be bad.  The interaction
between companies trying to make a profit and potential customers
deciding what to buy has led to all useful consumer innovation in
the past.  The socialist countries have made \F2no\F0 consumer innovations
but have ended up adopting almost every consumer innovation that
proved commercially successful in capitalist countries except
those beyond their organizational capacity such as checking accounts
or those leading to too much individual freedom such as consumer-available
copying machines.

	Assumption 5.  The proposed moratorium is feasible.
It seems to be assumed that unless the government finances an official
CIU pilot project, the utility will not come into existence.  In
my opinion, as I sit at my home terminal typing this, a prohibition-like
series of laws would be required to enforce Press's moratorium.
For example, Press and his Moratorium Enforcement Authority
don't even know that the Stanford Artificial Intelligence Laboratory
has an experimental news finder based on the Associated Press A-wire
and the \F1New York Times\F0 news service, and some people read
the news with it just because they want to know the news and not
with proper scientific intent.  Hundreds of lawyers can be employed
in arguing about whether a particular use of computers from someone's
home is an allowed business use or is a violation of the moratorium.

	It is remarkable that so much of the content of Press's article
is contained in unstated assumptions.  Now I would like to comment on a
point Press makes explicitly.


	He is against a federally financed experimental information
utility with "high inertia".  Considering what he means by it (and I
think he correctly expresses what the proponents of such an experimental
system mean), I also have my doubts.
\F1First\F0, such an official experiment would be a temporary local monopoly
aimed at a grander monopoly once it had been declared a success;
\F1second\F0, the social scientists dancing in attendance might
evaluate the benefits according to the current fashions among
social scientists rather than letting each user evaluate the
benefits for himself; and \F1third\F0, the issue is not whether
such services are ready for the public as a whole, but whether
the there is a segment of the public, however small, that perceives
benefits sufficient to pay what it costs to provide them the
services.  (\F1In this regard, an attempt is being made to form
a Bay Area Home Terminal Club, and I would welcome expressions of interest\F0).

	While I am not enthusiastic about an "official experiment", I doubt
that it probably would do any harm beyond wasting a little money.
I can't see Congress imposing Press's moratorium by law,
and it seems to me that the buying and selling of
home information services will start as soon as it is economical
to provide them, and this will be very soon.
On the other hand, some government support of the development
of information services technology would serve to keep the core
of this technology in the public domain and thus prevent monopoly.

	Press's concluding quote from E.M. Forster's short story, if taken
seriously
poses the issue of whether the woman of the future, unless
protected by a moratorium from overindulgence in
the Community Information Utility, is likely to become "a swaddled
lump of flesh about five feet high with a face as white as a
fungus", barely able to walk short distances, hold objects, or
breathe air.

	This article along with proposals to get ready for licensing computer
programmers strikes me as a manifestation of what I would like to call
the \F1bureaucratic ethic\F0.  Such proposals are also a manifestation of
the growing sense of self-importance of those members of the computer community with
a taste for committee meetings.

	In conclusion, let me confess that a large part of my motive in
disputing Press and also in conducting research aimed at home terminals is
that I would like to enjoy the benefits of widespread information services
before I reach the age of 72\.

			John McCarthy
			Computer Science Department
			Stanford, California 94305
cc: L. Press
acm.le3[let,jmc]:su-ai