UPDATE ON CARNEGIE-MELLON

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP88G01332R000300300003-3
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 4
• Document Creation Date: December 27, 2016
• Document Release Date: December 20, 2011
• Sequence Number: 3
• Publication Date: November 17, 1986
• Content Type: MEMO

Declassified in Part - Sanitized Copy Approved for Release 2011/12/20: CIA-RDP88GO1332R000300300003-3
,t
ROUTING AND RECORD SHEET
SUBJECT: (Optional)
A&t,i.1c.ca e I n.e Ug ence S. ct
FROM:
EXTENSION
NO.
DD A Registry
D
vicec
and FducatLon
ton os Tnarrn.
n
.
.
g
01 V%
? j
1026 CajC
.
DATE
28 Novembeh 1986
TO: (Officer designation, room number, and
building)
DATE
OFFICER'S
COMMENTS (Number each comment to show from whom
RECEIVED
FORWARDED
INITIALS
to whom. Draw a line across column after each comment.)
1. VVA
2 I]
C.1986
'IN
7D24 H2S
FYI at Catne
ie-Mellon
2.
/
g
ms~oH 03
DE(I79
3. Qaa
1989
. 5 nEC
198
ie
'~'~'
0 P'o )411;'~
5
1
4
.
~!
~j`a/ I s
.
O
I
5.
6.
7.
k
ljl~u~uu
f
I I
9.
10.
11.
12.
13.
14.
1S.
STAT
STAT
STAT
STAT
FORM 610 USEDI PREVITIONSOUS s u.5. Government Printing Offlos: 1555--454.$34$515$
1-79 E
Declassified in Part - Sanitized Copy Approved for Release 2011/12/20: CIA-RDP88GO1332R000300300003-3
 ............. .
Declassified in Part - Sanitized Copy Approved for Release 2011/12/20: CIA-RDP88GO1332R000300300003-3
Artificial Intelligence Staff
MISG/OIT
7:17 PM, Monday, 17 November 1986
Subject: Update on Carnegie-Mellon
The purpose of this note is to u date you on Carnegie-Mellon... i.e., what's
happening with it. and I visited Carnegie-Mellon last week to
discuss the one year's training in Al, per your (Ed) guidance. As an upfront
summary comment, the visit was quite successful and everything looks to be in
good shape.
and I met with three professors from the Computer Science Department,
including the Department Chairman, a Dr. Habermann. All three profs were quite
receptive to the idea of a special program of study, and they are flexible in
terms of custom-tailoring it to meet Agency needs. I might also add that all
three seemed to be "friendly" towards the Agency, so to speak. Here's some
relevant bullets from our discussions:
- I would work in the expert systems area, and a faculty advisor has been
tentatively identified. This looks to be particularly interesting as this would
plug me in as a working member of John McDermott's overall organization.
McDermott is well-known throughout the AI community as one of the top two or
three bonafide heavyweights in the field. (Among other things, he is the
creator of DEC's well-known XCON expert system.)
- I would also take a series of courses to broaden my background in the
underlying theory of AI. The Department Chairman suggested several CMU courses
which I made a note of and will plan to make a part of my program.
- They were all three insistent that I not start until the Fall Semester, '87.
CMU runs a series of familiarization/orientation/standardization courses then
(and only then) which all new students must take. (Actualy, the "Fall" starts
in August... for planning purposes, this means I would be leaving around 15
August or so.)
- Dr. Habermann suggested I visit CMU again for a day sometime in the Spring.
The purpose would be to do some advance planning with my faculty advisor. I'll
thus plan on doing that.
- CMU impressed me in a positive way as j* having a rather unique (for a
university) approach in that they emphasize the practical/hands-on as well as
and in coordination with the theoretical. In line with this, the Chairman
recommended I take a LISP course prior to my arrival so I will plan to do that
as well.
STAT
STAT
STAT
- As it turns out, Dr. Habermann was previously the Acting Director of CMU's
UNCLASSIFIED
ADMINISTRATIVE - INTERNAL USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2011/12/20: CIA-RDP88GO1332R000300300003-3
 Declassified in Part - Sanitized Copy Approved for Release 2011/12/20: CIA-RDP88GO1332R000300300003-3
ADMINISTRATIVE - INTERNAL USE ONLY
Software Engineering Institute (SEI), which is a DOD-sponsored activity designed
to improve overall software productivity. While SEI is technically separate
from the Computer Science Department, there is clearly a close and incestuous
relationship between them. The Chairman offered to broker introductions to
whomever I'd like to meet within SEI. I expressed strong interest in doing so
at the appropriate time.
- One of the professors I talked to is doing some interesting work in
"knowledge acquisition", which is a key element in building a successful expert
system. Coincidentally, I had met and talked with him several months ago at an
ORD-sponsored AI forum in Rosslyn. This little coincidence didn't hurt in the
least.
- By way of setting up a program, what seems to be a logical progression would
be to spend 6 months or so in the expert systems area, followed by 3 months in
the knowledge acquisition area and then to wrap it up with 3 months in the SEI
area. (I think in reality I'd end up starting off in the expert systems area
and then, after I got to see things first hand, adjust the program as
necessary...I want to be sure to keep the "practical" emphasis and not get
sucked into a purely academic mode.)
- The question of money was brought up by the Department Chairman pretty much
as you (Ed) had outlined it to me back in our intitial meeting on CMU. Dr.
Habermann stated that the true cost of educating a graduate student in CS at CMU
is approximately $33,000 per year. When corporations or the Government sends
them someone on a special program such as this, CMU's policy is to charge the
true costs at a minimum and it's customary to provide a fellowship as well. The
initial cost estimate given to over the phone was $75k. The
understanding we left with after talking in person was $35k "fellowship" funds,
plus the payment of normal tuition of approx $10,600.
STAT
- On the question of funds, I thought it best to get the dollars nailed down in
writing so that there are no misunderstandings come next Fall. I told Dr.
Habermann that the next step would be for the Agency to send Carnegie a letter
of understanding. After I returned, I then met wit to discuss this. STAT
He has kindly agreed to take this one for action, and to coordinate with OGC and
- With all of the foregoing in mind, I don't see any major problems which would
foul this up... everything really seems to have fallen into place. I'm most
excited about the program and, as I've said before, I feel most honored to have
been selected to represent the Agency at Carnegie-Mellon. Be that as it may, I
have a personal item to resolve that relates to my wife and some of the factors
that affect her by virtue of this whole matter...i.e., an employee couple type
thing. I don't want to get into the nitty-gritty here, but FYI I plan to get on
Gary's calendar (in your absence) to discuss this to see what the Agency might
be able to work out in her case (i.e., she has some concerns about what happens
to her recently renewed career as a DI Intelligence Assistant if she goes to
Pittsburgh with me, while being simultaneously unenthusiastic about the impact
on the family if I were to go to Pittsburgh by myself.) I'm hopeful this can be
resolved in some reasonable way, so I'm optimistically plugging ahead. --F I STAT
STAT
UNCLASSIFIED
ADMINISTRATIVE - INTERNAL USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2011/12/20: CIA-RDP88GO1332R000300300003-3
 Declassified in Part - Sanitized Copy Approved for Release 2011/12/20: CIA-RDP88GO1332R000300300003-3
STAT
UNCLASSIFIED
ADMINISTRATIVE - INTERNAL USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2011/12/20: CIA-RDP88GO1332R000300300003-3