R&D INSTITUTE
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
CIA-RDP79M00096A000500060021-0
Release Decision:
RIPPUB
Original Classification:
S
Document Page Count:
3
Document Creation Date:
December 16, 2016
Document Release Date:
October 19, 2004
Sequence Number:
21
Case Number:
Publication Date:
May 27, 1970
Content Type:
MF
File:
Attachment | Size |
---|---|
![]() | 178.98 KB |
Body:
Approved For Release 2004/10/27: CIA-RDP79M00096A000500060021-0
,..
27 May 1970
MEMORANDUM FOR:
SUBJECT R&D Institute
1. After reviewing the "Proposal for a Research and
Development Institute" (IHC-D-128/3), and listening to the
comments at the May 20 IHC meeting, I would like to offer
the following comments which you may find useful. Do with
them what you will.
2. There are two implicit assumptions in the proposal
which should be examined: (1) that the intelligence information
handling problems are somehow unique and different from those
encountered by any other organization, and are therefore
solvable only by unique methods, and (2) that there is a basis
of commonality in the problems faced by the intelligence community,
and therefore a concerted effort in attacking these problems is
somehow better than individual solutions. It is my contention
that both. of these assumptions are invalid.
3? With respect to the first, an examination of the lg
potential task. areas (listed in Attachment D to the Proposal)
reveals that, with the exception of the first task area which
is concerned with computer security controls and systems, there
are no unique areas of interest. Many of the areas use the words
"intelligence data" - whatever that is - in their descriptive
text. Tf, however, the word "intelligence" is omitted, what
are described are common problems plaguing any large organization
using computers, and these problems are being attacked-with vigor
in the outside world. For example, in the area of scientific
documentation, which covers roughly half of the proposed task
areas, the National Science Foundation Report #15 on Scientific
Documentation, lists 785 individual projects employing over 1~b0
people in 350 organizations. This work is sponsored by 125.
Societies, Foundations, Government Bureaus and Agencies. The
budget for this endeavor, roughly calculated at $50K per project,
amounts to $40M, or, if taken at $20K per investigator, to $30M.
~ ~ ~ -' _.
' is
Approved For Release 2004/10/17 `Cf`A= DP79M00096A000500060021-0
Approved For Release 2004/10/27: CIA-RDP79M00096~00500060021-0
~.~ET
Other task areas, such as the development of memories, the
design of distribution networks, etc., are certainly not
peculiar to intelligence agencies, and are being fully
investigated by Bell Labs (bubble memories), IBM, IT&T, and
numerous other commercial firms who have a high monetary
interest in their outcome. Many of the commercially available
programs, hardware and techniques are not being exploited to
their fullest extent, and indeed, may not be known to the
individual agencies. In view of the enormous effort that is
being expended to solve problems in the task areas cited, it
is not very likely that any contribution, whether large or
small by agency or community standards, will have any appreciable
impact or hasten the development of solutions.
4. The second assumption appears valid only if not
examined too closely, and was the basis for the COINS concept.
It seemed logical to question why each Agency should maintain
separate biographic files, for example,-and an attempt has
been made to consolidate. these. for community use. Disregarding
the mechanical problems of whether to use separate computers
or a central unit, what languages to use, how to switch queries
and responses, and querying and security problems, (which are
not trivial.), it turns out that each Agency creates and main-
tains their own biographic files to best serve their own require-
ments. For this reason, these files are, in each Agency, widely
different in method of storage, content, structure and query
method. The lack of common purpose, except in the broadest
possible sense, lack of common data bases and lack of common
equipment therefore precludes even the definition of community
problems, much less their solution. Moreover, the effort to
create a common data base to serve these divergent interests
seems to us almost certain to increase the cost of file maintenance
rather than reduce it.
5. The ADP management structure witYiin the CIA (and presum-
ably in the other intelligence agencies) is highly decentralized,
which tends to encourage local initiative. This is probably a
good thing; however, it also reduces the long range planning
e~'fort, gives inadequate project control and does permit duplica-
tion of effort - all of which. are necessary on an Agency and
community wide basis to properly set up an R&D Institute which
will produce useful output. This does nat mean that a highly
centralized system is to be preferred to the present loose,
decentralized system. A community wide, centralized R&D effort
would be so heavily bureaucratic and so remote-from the user
that there would be no advantages to be gained.
~~'~~
.:~:; : ~.
Approved For Release 2004/10/27: _C~A P79M00096A000500060021-0
Approved For Release 2004/10/27: CIA-RDP79M00096A000500060021-0
6. What is needed, then, is some form of centralized
review and planning mechanism which would permit local initia-
tive to flourish and at the same time prevent the technical
wheel from being reinvented, or, worse still, square wheels
from being produced. Whether the mechanism is a full time
inter-agency staff, or a technological gatekeeper group from
one agency with a charter to make community wide recommendations
to higher authority, or some other mechanism, is a different
question and beyond the scope of this paper. Indeed, the ASPIN
Committee has been examining many of these points within the
CIA for a year and will undoubtedly arrive at some conclusions
regarding centralized versus decentralized management systems.
Perhaps a similar study on Community Systems, covering such
concepts as COINS, VIDEOFILE, the R&D Institute
and others would provide a con o e magnitude of work
currently underway and give fresh insight into ways to coordinate
these efforts without completely stifling them., ,,r
25X1
25X1
~~'!~~
_3_
Approved For Release 2004/10/27: CIA-RDP79M00096A000500060021-0