COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\\M0BASL30\M1BASI30\.
C00016 ENDMK
C⊗;
\\M0BASL30;\M1BASI30;\.
\F0


Dear Dr. Morgan:

\J	Thanks for sending  the detailed commentary on  our proposal.
We  do plan  another  proposal.   I  think we  can meet  some  of the
objections  of  the   reviewers  but  not  all.     Let  me   proceed
systematically through your letter.

	1.  The reviewers  are correct  in their  criticism that  the
proposal  is not as concrete as it should be.   I had hoped to get by
with few details  for two  reasons: First, many  of the services  and
experimental  home  terminal system  should  provide  will depend  on
getting access to  other people's computer  systems, finding  whether
airline or play reservations are available, and negotiating access to
these systems  is a major job  which I hoped to  undertake as part of
the project rather than before it.  Getting and keeping our access to
the  Associated Press  wire has  involved about  ten letters  and two
trips  to New York,  and I fear  that some of  the other negotiations
will be just as  lengthy.  I guess there  is no choice, but to  do at
least some of this work on speculation.  The second problem is that I
had hoped to be able  to adapt the projects  to be undertaken to  the
interests and  ideas of  the graduate  students and programmers  that
would be employed on the project.  We can certainly give this up.

	The  omission  of  message  sending  in  the proposal  was  a
mistake.   I have  found  the ARPA  network message  facilities  very
useful  as  well as  the  message  facilities  in our  own  computer.
However,  it is a bit more difficult to  see how much benefit a small
randomly self selected  group of home  terminal subscribers will  get
from message facilities.

	I can treat the costs  of home terminals in the new proposal.
I think it  will be possible  to make $500  terminals, and that  this
price will  come about  in five  years without  a special project  on
behalf of home terminals.

	2. Several of the reviewers take the proposal to task for not
involving social scientists.   The first reviewer  even says, "But  I
believe that  fundamentally the major  issue is one of  economics and
one  of  changing  existing  social  structures  to  allow  this  new
revolutionary development."  I  think this  reviewer  is mistaken  in
supposing  that social structures  have to  be changed to  allow home
terminals; home terminals  will be  purchased in  the present  social
structure if the price is right and the services are worthwhile.  The
widespread  use of home terminals  will then cause  changes in social
structure.

	I am not very enthusiastic about involving  social scientists
but would be happy to do it if NSF is willing to pay for it.  I don't
know what  question they would investigate.  The simple futurology of
home terminals has been done by the Institute for  the Future, and it
didn't seem to me that anything clear but non-obvious came of it.

	3. Here are some remarks concerning the budget:

	  a. The  main computer  resources the  project will  use are
computer time and disk space, and moreover our low speed  multiplexor
is inadequate to add a number of external terminals.  However, we are
not a cost center and are not allowed to charge directly for computer
time.   Therefore, we  proposed hardware  that,  in combination  with
money we  hope to get from  other sources, will enable  us to improve
the  system to where  it can  handle the additional  load coming from
this and other projects.

	  b.  While we  do  not  propose  to  develop  the  terminals
themselves, we need  some in order to have  some customers.  Suitable
terminals today are quite expensive.

	 c. Leaving out the biographies of Earnest and Winograd was a
goof.

	4. Concerning the analysis of the effects on users.

	In my opinion, not much can be done beyond measuring how much
they  use the system,  what services are  used how  much, and whether
they are willing to pay for the services.  This is because effects on
users of a  pilot project will be minor until  a system becomes quite
widespread.

	I suppose this view will worry some reviewers, because if you
can't measure the effects until the system is widespread, what if the
effect is bad.   I don't consider this much  of a problem, because if
the users don't like the system, they  won't use it.  All we need  to
prejudge is that home terminals are not like  heroin, bad for you but
addicting.

	5.  The  third   reviewer  doesn't  see  the  point  of  home
terminals.   He thinks  that  being able  to publish  something  with
trivial effort  is trivial. I  thought that the Home  Terminals paper
went  into that,  and I can't  tell what else  to say  to him without
being able to question him.

	He also wants an  elaborate paper study about the  long range
social implications  of home terminals.   It seems to me  that such a
study would be as speculative as my original paper, and I don't  want
to take part in  it. Perhaps I am being unimaginative  here, and some
social scientist could convince me otherwise.

	6. Naturally I like the fourth reviewer, because he likes our
proposal. His worry that making everything available would lead to an
overload is mistaken.  If readers today found their reading matter by
going  through the card catalog of  the Library of Congress seriatim,
they would also be overloaded.   People's need for help would  make a
market for  periodical documents in  the system pointing  to what the
reviewers thought  worthwhile. However,  the system  itself need  not
have  an official  opinion  about what  is  important, and  different
"reviewing journals" would  have different opinions. I agree with his
concern that efforts  in privacy  and security not  become an  effort
sink, and don't plan to put a substantial fraction of our effort into
it.

	7. The  fourth reviewer thinks that all the applications will
be developed as  the demand  develops.  In  my opinion, a  systematic
effort is  required to reach a collection  of applications sufficient
to justify  a user  getting a  terminal.   He also  finds this  topic
inappropriate for  an AI laboratory.  Our interests are  broader than
AI.

	He also  questions whether the work  will have impact outside
the AI Lab.  I think our A.P. based news service already has had such
impact and we haven't  even published a paper about it  yet.  I think
publication  of the results of our  experiments together with letting
people log in from afar and experiment with the services will have an
impact.

	8. We have discussed  the possibility of a joint project with
TYMSHARE, a large national time-sharing service organization  located
near us  and having  a number  of PDP-10s.   They  are interested  in
collaborating with  us and are considering  putting in some resources
of their own.  Would this make the project more attractive to NSF?

	We will make another proposal unless it seems hopeless.  Here
are some considerations:

	1. We will  not propose a pure paper study  of the effects of
home  terminals.   As far as  I am concerned,  the results  of such a
study are incorporated in the Computer Terminals in the Home paper.

	2.  The  minimal proposal  would  be  to  develop  some  home
terminal  applications and test  them in  the lab  without installing
terminals in homes. This would allow quite a small proposal,  whereas
installing terminals  in homes  is fairly  expensive and perhaps  not
worthwhile   unless  a  substantial  number   of  services  are  made
available.  We  have already developed  one more application,  namely
book reading (someone  gave us a mag tape  of \F1Wuthering Heights\F0
and we  have just finished but not tested a program for reading books
on a display terminal and remembering one's place.

	3. An intermediate level would be to make some terminals with
suitable  services  available in  the  Stanford  Library  and in  the
Student Union.   We can  propose this  if we can  think of  a set  of
services useful in this context. This would be relatively inexpensive
since we  could expect to learn more from a  few terminals than if we
put them in homes.

	4.  I  will  send  you  another  letter  with  some  specific
applications that  we might work  on shortly.   However, I  will also
phone you shortly to get your reaction to this letter.