COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	Combination of several files from ess,jmc
C00012 ENDMK
C⊗;
Combination of several files from ess,jmc

ad.net[ess,jmc]
SU-AI


Artificial Intelligence Project
Computer Science Department
Stanford University
Stanford, California 94305
PDP-10 and PDP-6

I. Personnel

Director - John McCarthy

Executive Officer - Lester Earnest	x4202
Associate Director- Jerome Feldman	x4532
NIC Station Agent - Barbara Barnett	x2800
Network Software - Andy Moorer		x4971	JAM

II. Installation Type

	The computer system includes PDP-10 and PDP-6 processors,
256K of core, a swapping disk, an IBM 3330 for file storage including
a user disk pack, IBM compatible 7 channel tape, Dectape, about 40
keyboard-display terminals, and i-o equipment for robotics research
including TV cameras, arms, and a cart.  The system software is a
modified version of the D.E.C. system for the PDP-10.  The major
languages used are the FAIL symbolic assembler, the SAIL algebraic
compiler, LISP, and micro-PLANNER.


VI. Interests and capabilities.

	The Stanford Artificial Intelligence Laboratory has potential
interest in all areas of AI and mathematical theory of computation.  The
following activities a receiving significant effort at present:

	1. Computer vision.  Edge finding, body and scene description,
use of color, parallax, and controlled illumination.  If and when the
it is feasible to transmit pictures over the network, network collaboration
in this area may become feasible.
The leaders of this work are Jerry Feldman (JAF) and Tom Binford (TOB).

	2. Robotics.  Projects in arm control and control of a vehicle
are in progress.  Feldman and Binford are in charge here too.

	3. Mathematical Theory of Computation.  The group in this area
includes John McCarthy, Robin Milner, Richard Weyhrauch, Ralph London,
Shigeru Igarashi, and David Luckham and Whitfield Diffie.  Two programs
for checking mathematical proofs may be of interest to network users.
The first checks proofs in predicate calculus, and Diffie (WD) is
responsible for it.  The second which checks proofs in version of
Scott's logic is in the charge of Milner (RGM) and Weyhrauch (RWW).
Both programs have been used for checking the proofs of correctness
of computer programs.  Network use of these programs (at least
experimentally) is probably quite feasible.

	4. Computer theorem proving.  Luckham (DCL) has a resolution
theorem prover of considerable power as such things go.  Its experimental
network 
use is probably also feasible.

	5. Arthur Samuel (ALS) heads our speech recognition work.
There may be some network interest in our speech programs.

	6. There is a project for a new version of LISP called LISP 70.
Experimental versions will be available in the summer of 1972, and
information may be obtained from Horace Enea (HJE), Dave Smith (DCS)
or Larry Tesler (TES).

	7. Yorick Wilks (YAW) is writing programs for translating
between English and French.

	Apart from the research work in AI and MTC, there may be some
interest in our system software.  Specifically, other installations
may be interested in experimenting with the FAIL assembler which is
very fast, the SAIL compiler which is ALGOL + LEAP structures, or
our editor SOS.  Inquiries about documentation should be directed to
Barbara Barnett (BB) or Marilyn Mullins (MLM).


adfect[ess,jmc]
Adfected [A specialized form of affected].  Compounded.  Of
equatons in algebra: Containing different powers of an
unknown quantity.
1695 Allingham Geom. epit. "The method of finding the root
of an adfected equation".  1728 Campbell in Phil. Trans.
"Every adfected quadratick equation ax: - Bx + A = 0 whose
Roots are real."  1870 Todhunter, Algebra.  "Quadratic
equations which contain the first power of the unknown
quantity as well as the square are called adfected quadratics."

(The : after  a  in the equation may be ! or may be absent.)

all[ess,jmc] list of essays as of april 1973
CAR.ESS,
CHICAR.ESS,
EDUCAT.ESS,
ENERGY.ESS,
CRIME.ESS,
DELIV.ESS,
DIG.ESS,
EARLID.ESS,
HOTER.ESS,
INTRO.ESS,
LONG.ESS,
MONOP.ESS,
OBJEC.ESS,
PLANE.ESS,
SHOCK.ESS,
SLOGAN.ESS,
SOCIAL.ESS,
SPACE.ESS,
STYLE.ESS,
TCLUB.ESS,
TECHNO.ESS,
TECIND.ESS,
UNDER.ESS,
WOMEN.ESS


ess.msg[ess,jmc]
∂26-May-80  1806	JMC  
Increase feasible commuting distance as goal of technology.


puzz[ess,jmc]
Here is the solution to your puzzle of last night:  Let  p(n) be the
position  mod n  of the person who should be in the  n th place and
let  r(n)  be the amount the table must be rotated to bring him to
his correct place - all assuming  N  people numbered  0,1,...,N-1.
Unless the  N  r(n)'s  are precisely the  N  numbers  0,1,...,N-1,
rotating by the missing number will bring everyone to the wrong place.
If they are, then adding up the  N  congruences  p(n) + r(n) ≡ n (mod N),
each expressing that rotating the table by  r(n)  will bring the
person who should be at position  n  from position  p(n) to that position,
gives the congruence  (N↑2-N)/2 + (N↑2-N)/2 ≡ (N↑2-N)/2 mod N.
since the first summands, the second summands and the right hand sides
each form the set {0,1,...,N-1}.  Setting  N = 2K gives a left side
congruent to zero and a right side congruent to K.  This shows that
the  r(n)'s  cannot be all different, and so there must be a rotation
putting everyone in the wrong place.

	If  N  is odd, then setting  p(n) = -n, i.e. reversing the
order of the guests produces a configuration in which any rotation
puts some guest in the right position, because then  r(n) = 2n, and
since  2  and  N  are relatively prime, the  2n  are again a complete
set of residues.  The only remaining question is what other suitable
configurations exist in the odd case.


serial[ess,jmc]
Serial numbers
superphone 7700, model 7710, serial 007377
Vera's camera	644430
Vera's lens	379453

Telephone service
as of December 1978

unlimited local			5.70
one hour calling bay area	3.45
	additonal .0575 per min
3 extensions			2.30
2 trimline			2.50
1 touchtone			1.20
3 touchtone phone		0.75

total				15.90

bank accounts
64817 90156	Vera house
64817 90261	Vera estate