perm filename FEB.ME[LET,JMC] blob
sn#089170 filedate 1974-02-28 generic text, type T, neo UTF8
∂28-FEB-74 2137 1,MM
do you have opinion about desirability of arpa ai committee.
i do
∂27-FEB-74 1658 S,REG
I WAS DISAPPOINTED BY THE RESULTS OF THE ARPA MEETING. I THINK
FIELDS IS UNRECEPTIVE TO OUR NEEDS AS STATED IN OUR POSITION
PAPER. ONE THING THAT HE DID SUGGEST WAS THAT WE ATTEND T
A FRONT-END CONFERENCE (MARCH 14,15 IN DC) TO PRESENT AGAIN
OUR DISPLAY SERVICE IDEAS.
I'VE BEEN TO DEC. THEY APPEARED VERY HAPPY TO SEE US, AND VERY
CORDIAL. THEY'VE DECIDED THAT THEY NEED A TREE STRUCTURED FILE
SYSTEM; THERE'S A STRONG POSSIBILITY THAT IF WE COOPERATE WITH THEM
IN MAKING A SPECIFICATION, THAT THEY'LL DO A SUITABLE IMPLEMENTATION.
THEY DON'T APPEAR TO BE WILDLY INTERESTED IN DISPLAY STUFF AT THE MOMENT,
BUT I THINK THAT IF WE DID A GOOD JOB PUTTING DISPLAY STUFF IN
VIROS THAT THEY'D USE IT.
WILHELM GAVE ME PERMISSION TO TAKE A SET OF VIROS SOURCES, BUT
PROBABLY WE'LL HAVE TO WAIT FOR RPH TO SEND THEM.
PETE CONKLIN OF DEC ALSO EXPRESSED INTEREST IN A SUCCESSOR
TO PUB. THEY'RE HAVING SIMILAR PROBLEMS DOING DOCUMENTATION,
WE MIGHT JOINTLY COOK UP A SPECIFICATION AND HAVE EITHER ONE
OF US (SAIL OR DEC) DO THE IMPLEMENTATION.
ALSO, LICKLIDER SAID THAT WE SHOULD NOT FEEL THAT WE WERE BEING FORCED
TO TENEX. HE SAYS THAT OUR BUSINESS IS RESEARCH AND THAT IF TENEX IS
INIMICAL TO THAT RESEARCH, WE SHOULD NOT USE IT. I BELIEVE
THAT HE COULD BE TALKED INTO LETTING US DO OUR OWN SYSTEM IF WE COULD
SHOW THAT 1. IT WAS A RESARCH GOAL IN ITSELF, AND 2. IT WOULD
ULTIMATLY PROVIDE A STRONGER RESEARCH TOOL FOR OTHER WORK, AND
3. THEdTRANSIENT NOT BE TOO BIG.
SINCE I'VE TALKED TO DEC, I'M MORE ENTUSIASTIC ABOUT TRYING VIROS.
∂26-FEB-74 1242 PC,RWW
Do you know where I can find LCOM0...n and any documentation there is
for them? Ralph London's memo on correctness mentions a write-up by
you. Bruce
∂25-FEB-74 2345 network site CASE
-------
Date: 26-FEB-74 0245-EDT
From: CROCKER at CASE-10
Re: AI Review
cc: Newell at CMU-10A, PHW at MIT-AI, Minsky at MIT-AI,
MM at MIT-AI, HART at SRI-AI, Crocker at ISI
- - - -
I have talked with most of you about the possibility of a high level
review of A.I. that I would like to crank up pretty soon, and it
is now time to start serious action.
I have two problems I want to solve.
1. While at ARPA, I have been a pretty firm defender of A. I. and
have not probed any of the scientific issues very deeply myself.
I am basically uncomfortable with this position, for my representation
of the field is necessarily faulty. I would like to have a much better
understanding of just how we stand with respect to the classic
goals. This desire is much more than an attempt to satisfy ARPA
management that everything is being run tightly -- though it
should have that effect too -- but rather an attempt to solidify
my position so I can operate from a position of greater strength.
2. With the exception of the A.I., Automatic Programming and
Speech Understanding systems research, things are pretty dull
on the computer science side of our operation. (The packet
switching technology is pretty exciting, but it's a quite different
game.) In between the communication science and the A.I. type
computer science we are doing a lot of systems building. The systems
building is fun and extremely useful, but it's much too easy and
holds very little real content. What's needed are some new research
projects which push at the boundary somewhere. I am particularly
interested in A.I. application type projects, along the lines
of the Speech Understanding systems effort, but I am open to other
ideas.
I have thought of a specific mechanism for approaching these
problems, which I will shortly discourse on, but the main point is that
I want to work on the above two problems and am open for advice at
all levels of discussion.
The specific mechanism:
A committee to be formed pretty soon, about a dozen scientists, about
2/3 from the ARPA A.I. community with the rest not -- either not ARPA
or not A.I, or not both -- but maybe the proportion should be
reversed, to take a deep look at the 4 labs by site visits and by
sessions by themselves and with larger groups.
To take place during the two week period April 22 and following.
To be preceded by papers from the labs which delineate the goals,
the methods of attack, the probable timeframes, the approopiate measures
of success, the past accomplishments.
The committee responsible for a very critical look at the quality
of the science and the prognosis for success. [If it's a 50 year
venture, then I want to know it. Question, can we tell the difference
between a 25 year venture and a 50 year venture??] Perhaps
experienced scientists not steeped in A.I. but well-grounded in the
philosophical and mathematical prerequisites and broadly experienced
in other scientific matters would help keep the whole probe calibrated.
[Who might such a person be?]
The committee responsible for making recommendations of likely
topics for separately funded and managed research in collateral
areas.
I have thought some about who I would like to see on the committee,
but I expect the composition to be determined generally by
consensus or negotiation. I have carried on a fairly
specific set of discussions with Michie, and will have to invite him
formally pretty quickly if he is to participate, but that is entirely
because of the formalities he has to go through to make arrangements
to come here at the time I suggested.
Before we continue to build the committee, I'd like to have
a short dialog on the ground rules for this effort and get some
feedback from you about whether this looks like an effective mechanism
for solving my problems or whether there might be some other
ways to do business. Let me know pretty quickly -- say this month --
what your immediate reactions are.
Please respond to Crocker@ISI.
Thanks,
Steve
-------
∂18-FEB-74 1847 R,AJT
JOHN:
I've put in your mailbox a document which is my second attempt
at a thesis proposal. I hope you will find it more interesting/less
vague than the previous one. I have talked it over with Richard W, and
he seems to think it's OK. What do you think?
I shall be away until the end of the week: perhaps we can chat about it
when I get back.
Thanks. Arthur
∂17-FEB-74 2045 1,DEW
I borrowed Botvinnik. Ta, Dave.
∂16-FEB-74 1155 1,REM
RE THE INDEX TO LISP FUNCTIONS YOU WANTED.
DOES THERE ALREADY EXIST AN INDEX YOU WANT ME TO ADD TO,
OR DO YOU WANT ME TO DESIGN ONE FROM SCRATCH?
I PROPOSE (1) A SPECIFIC INDEX TO VERY-USEFUL FUNCTIONS (2) A COLLECTIVE
INDEX TO PACKAGES OF FUNCTIONS THAT ARE SPECIAL-PURPOSE:
(PACK A1 A2 ... AN) LMMSAV[1,REM] PACK ATOMS INTO ONE ATOM
<PACKAGE> HASH.LSP[1,REM] ROUTINES TO HASH S-EXPR'S AND ATOMS
∂15-FEB-74 1526 1,KNM
i had to go back to the sadistic doctor to finish my foot operation. the
comprehensive exam info is on comp.s74 but i need to work on the formating
of bibliography. could you please try to read over at least some of the
application folders over the weekend as they are hollering loudly for them
back. sarah said you were going skiing lucky. have a grand time and may
the snow gods smile on you.-kasee
∂01-FEB-74 1244 D,AJT
John: I should be delighted to talk to you at any length about what I'm
doing and intend doing, anytime that you have time to spare. I realize
that you're very busy, but I regard you as my chief mentor, so I get very
diisturbed when you seem dissatisfied with my stuff. Sorry to be a bother.
Arthur.
∂06-FEB-74 1254 H,RF
YOU MAY BE INTERESTED IN COMMUT.FOL, SETCOM.FOL, AND THEORM.FOL ON [MAP,RF].
∂05-FEB-74 0944 FOO,DBA
Just heard I didn't get the Essex job - Pat Hayes says they decided they
really wanted a commercial DP person if possible, and they got one.
Thanks for refereeing. Bruce.
∂04-FEB-74 0951 ACT,REG
TENEX-132 will be available from BBN on 25 FEB (coincides
with the ARPA/TENEX meeting.
∂03-FEB-74 1840 DOC,JEG
I am TGR and will finish my Ph.D. by the end of the quarter.
∂10-FEB-74 0956 network site AI
- - - - - - - - - - - -
02/10/74 12:48:52
From: MINSKY@MIT-AI
- - - - - - - - - - - -
This may be a duplicate message. Thanks for summary of meeting.
I do want to come out for a few days before very long.
Any suggestion about good time slot?
Frame paper about finished and much better. Would you care to read it?
∂09-FEB-74 2300 11,BO
I think I fixed it. Added two local commands for the IMLAC--
<ctl><esc>E to allow the special features that hung you up,
<ctl><esc>N to disable them. You can try them and see whether
they work. Say RU FIXIML[IML,BO]. When it types a *, say
<ctl><esc>B and <ctl><top><shift>S. It will reload with
the new version. If it doesn't work, the old FIXIML is still
on the system.
∂10-FEB-74 1537 HMF,YAW
HMMMM...IVE NEVER SEEN THAT ONE BEFORE! SORRY ABOUT THAT---MAYBE YOURE PROVING
THAT THERES NO LANGUAGE SYSTEMS THAT ONE CAN JUST BE TURNED LOOSE ON---THOUGH
IT DOESNT USUALLY LET PEOPLE DOWN SO RAPIDLY AS THAT.ID BE GLAD TO KNOW WHAT THE
SENTENCE WAS THAT DID IT.
∂12-FEB-74 1028 1,KNM
the latest crank letter enclosed a picture of himself--certainly keeps the mail
interesting
∂12-FEB-74 1021 1,KNM AT TTY36 1021
ann hardy phoned. they cannot make it friday after all, but will come tues 19th
at 3--kasee
∂12-FEB-74 1017 network site SRAI
-------
Date: 12-FEB-74 1012
From: WALDINGER at SRI-AI
Re: A AND C.
- - - -
WE WERE MERELY PERPETUATING AN ERROR INTRODUCED BY JIM KING IN
HIS THESIS. THE VERSION IN THE AI JOURNAL WILL HAVE A NOTE
ABOUT THE DISCREPANCY.
RICHARD WALDINGER
-------
∂11-FEB-74 1155 1,KNM
p.s. on that letter for the folks writing in re chronicle. after a search thru
your files i could find no people.le1. i located people[226,jmc] which i decided
couldn't believe was it. after i mashed thru directories i came up with public.le1
which i guess is what you mean't.
∂11-FEB-74 1140 IML,BO
My IMLAC fix worked for me. I got into BBN with apparent success from Polya Hall.
The thin? to do i∂ tell the/IMLAC <ctl><esc> N before you run TELNET, to disable
the command features.
∂11-FEB-74 1100 1,KNM
1. i told ralph about blundering into xgp,sys and how i tried to delete without
success last december, so i am properly scathed as he yelled at me at the time.
2. i have sent people to the paranoic from oakland, ms. alexander. do you want
one to go to the bloke who offered to help you design "humanly intelligent"
machines?
3. crime and monop sent out to kaplan.
4. ershov telegram sent.
5. i have several student application folders you are suppose to review asap.
6. busy week--remember court and crocker on wednesday, a visiting prof.(kochen)
wants to see you--i told him 3 on thursday, let me know if not ok, friday you
have comp exam meeting at 11 at serra and tymshare at 2 here.
7. speaking of the comp exam committee--i need that solution pronto. luckham is
still holding out despite omnipresent pestering, but i shall keep at it. i
think it important to get the solutions out as guynn is constantly besieged by
the young einsteins panting for the true word and is beginning to pull her hair
out and my ears off.
8. although you are now a celebrity and have many calls upon your time, sometime
we should sandwich an hour in for some mundane chores like cleaning up the
mail.
9. hope you are all recovered from your bout with the nasties. i had a touch
friday and spent a most unpleasant day alternating between bed and bathroom.
someone should undertake seriously purging the flu bugs from the list as i
believe they are overtaking the common cold as most irritating.
10. looking forward to your tube performance--which will account for my droopy
eyes on tues morn. kasee