/LMAR=0/XLINE=3/FONT#0=BASL30/FONT#1=BASI30/FONT#2=BASB30/FONT#10=BAXM30/FONT#11=ZERO30/FONT#5=MICR25/FONT#6=NGB30
␈↓ ↓H␈↓¬␈↓ α≡OFFICE MEMORANDUM  *  STANFORD UNIVERSITY  *  OFFICE MEMORANDUM

␈↓ ↓H␈↓ε␈↓ β=STANFORD LOW OVERHEAD TIME-SHARING SYSTEM

␈↓ ↓H␈↓α␈↓ ¬]October 18, 1978




␈↓ ↓H␈↓αTo:␈↓ αH␈↓William F. Miller, Provost

␈↓ ↓H␈↓␈↓αFrom:␈↓ αH␈↓John McCarthy, Director of LOTS

␈↓ ↓H␈↓αSubject:␈↓ αH␈↓Opinions

␈↓ ↓H␈↓Dear Bill:

␈↓ ↓H␈↓        Thank␈α⊃you␈α⊂for␈α⊃your␈α⊃kind␈α⊂memo.␈α⊃ Some␈α⊃of␈α⊂the␈α⊃events␈α⊃that␈α⊂prompted␈α⊃my␈α⊃declining␈α⊂the
␈↓ ↓H␈↓invitation␈α∞to␈α∞appear␈α∞before␈α∞the␈α∞study␈α
committee␈α∞turned␈α∞out␈α∞to␈α∞be␈α∞unfortunate␈α∞coincidences,␈α
and
␈↓ ↓H␈↓anyway␈αI␈αdecided␈αto␈αassert␈αmy␈αpoint␈αof␈αview␈αrather␈αthan␈αjust␈αwithdraw,␈αso␈αI␈αwill␈αcome.␈α However,
␈↓ ↓H␈↓I should summarize my views and tell you what made me unhappy and angry.

␈↓ ↓H␈↓        I have long believed

␈↓ ↓H␈↓        1.␈α∂The␈α∂best␈α∞way␈α∂to␈α∂compute␈α∂is␈α∞interactively,␈α∂and␈α∂I␈α∂invented␈α∞the␈α∂concept␈α∂of␈α∂general␈α∞time-
␈↓ ↓H␈↓sharing␈α
and␈α
initiated␈α
the␈α≡rst␈α
work␈α
on␈α
it␈α
when␈αI␈α
was␈α
at␈α
M.I.T.␈α I␈α
also␈α
initiated␈α
the␈α
≡rst␈αdisplay-
␈↓ ↓H␈↓based␈α∩time-sharing␈α∩system␈α∩on␈α∩the␈α∩PDP-1␈α∪at␈α∩Stanford␈α∩and␈α∩the␈α∩≡rst␈α∩system␈α∩based␈α∪on␈α∩having
␈↓ ↓H␈↓terminals␈αin␈αall␈αo≠ces␈αat␈αthe␈αAI␈αLab.␈α I␈αhave␈αalways␈αbeen␈αdisappointed␈αthat␈αthis␈αhas␈αnever␈αbeen
␈↓ ↓H␈↓considered to qualify me to take part in Stanford's computer planning.

␈↓ ↓H␈↓        2.␈αThe␈αuse␈αof␈αinteractive␈αand␈αmost␈αother␈αkinds␈αof␈αcomputing␈αare␈αstill␈αsubstantially␈αinhibited
␈↓ ↓H␈↓by cost.

␈↓ ↓H␈↓        3.␈αIt␈αis␈αpossible␈αto␈αorganize␈αa␈αcomputing␈αoperation␈αserving␈αa␈αlarge␈αuser␈αpopulation␈αand␈αstill
␈↓ ↓H␈↓keep␈α
personnel␈αcosts␈α
well␈αbelow␈α
the␈α
hardware␈αcosts.␈α
 I␈αbelieve␈α
that␈α
LOTS␈αwith␈α
a␈αpaid␈α
sta≥␈αof␈α
four
␈↓ ↓H␈↓and␈α
about␈α
3000␈α
di≥erent␈α
users␈α
each␈α
quarter␈α
has␈α
demonstrated␈α
this.␈α
 While␈α
I␈α
underestimated␈αthe
␈↓ ↓H␈↓amount␈αof␈αcomputing␈αpower␈αrequired␈αto␈αswitch␈αstudent␈αuse␈αto␈αan␈αinteractive␈αmode,␈αwe␈αare␈αgetting
␈↓ ↓H␈↓by.␈α
 Moreover,␈α
the␈α
main␈α
weakness␈α
of␈α
LOTS␈α
is␈α
not␈α
enough␈α
computer␈α
rather␈α
than␈α∞an␈α
insu≠cient
␈↓ ↓H␈↓sta≥.

␈↓ ↓H␈↓        Let␈αme␈α
assure␈αyou␈α
that␈αI␈αhave␈α
not␈αbeen␈α
a␈α≡gurehead␈αat␈α
LOTS␈αeither.␈α
 I␈αhave␈αinitiated␈α
most
␈↓ ↓H␈↓of␈αthe␈αmajor␈αactivities,␈αand␈αI␈αhave␈αmade␈αthe␈αmajor␈αpolicy␈αdecisions.␈α Ralph␈αhas␈αdone␈αan␈αexcellent
␈↓ ↓H␈↓job,␈α
but␈α
if␈α
he␈α
leaves␈α
for␈α
a␈α
job␈α
with␈α
greater␈α
responsibility,␈α
we␈α
will␈α
be␈α
able␈α
to␈α
replace␈α
him.␈α (We␈α
may
␈↓ ↓H␈↓need another person if we can't get someone as versatile).

␈↓ ↓H␈↓        4.␈α∞Stanford␈α
could␈α∞meet␈α∞its␈α
general␈α∞computing␈α
needs␈α∞with␈α∞many␈α
fewer␈α∞people␈α∞than␈α
operate
␈↓ ↓H␈↓SCIP.␈α Moreover,␈αit␈αcould␈αhave␈αswitched␈αits␈αoperations␈αto␈αinteractive␈αcomputing␈αmuch␈αfaster␈αthan
␈↓ ↓H␈↓it did.

␈↓ ↓H␈↓        5.␈α
While␈α
there␈α
is␈α
nothing␈α∞that␈α
intrinsically␈α
prevents␈α
a␈α
centralized␈α
organization␈α∞from␈α
being
␈↓ ↓H␈↓e≠cient,␈α∩centralizing␈α∩removes␈α⊃some␈α∩of␈α∩the␈α⊃inhibitions␈α∩against␈α∩organizational␈α∩in∨ation.␈α⊃ When
␈↓ ↓H␈↓engineers,␈α⊂physicists␈α⊃and␈α⊂computer␈α⊂scientists␈α⊃provide␈α⊂for␈α⊂their␈α⊃own␈α⊂computing␈α⊂needs,␈α⊃they␈α⊂are
␈↓ ↓H␈↓unlikely␈αto␈αoversta≥␈α
computing␈αat␈αthe␈α
cost␈αof␈αhiring␈α
engineers,␈αphysicists␈αand␈α
computer␈αscientists.
␈↓ ↓H␈↓Given␈α⊗the␈α⊗bureaucratic␈α⊗ethos␈α⊗of␈α⊗the␈α⊗1960s␈α⊗and␈α⊗1970s,␈α⊗it␈α⊗would␈α⊗have␈α⊗taken␈α⊗very␈α∃strong
␈↓ ↓H␈↓management for SCIP to have escaped organizational in∨ation.
␈↓ ↓H␈↓αWilliam F. Miller␈↓ ¬]October 18, 1978␈↓ 
nPage 2␈↓ 


␈↓ ↓H␈↓        While␈α⊃decentralization␈α⊃would␈α⊃probably␈α∩make␈α⊃scienti≡c␈α⊃and␈α⊃educational␈α∩computing␈α⊃more
␈↓ ↓H␈↓e≠cient,␈α∩di≥erent␈α∪methods␈α∩may␈α∩be␈α∪required␈α∩to␈α∩reduce␈α∪the␈α∩cost␈α∩of␈α∪administrative␈α∩computing.
␈↓ ↓H␈↓Maybe␈α∞combining␈α∞decentralization,␈α∞the␈α∞use␈α∂of␈α∞outside␈α∞services␈α∞and␈α∞across-the-board␈α∂budget␈α∞cuts
␈↓ ↓H␈↓would work.

␈↓ ↓H␈↓        These␈αgeneral␈αviews␈αare␈αnot␈αbased␈αon␈αexperience␈αwith␈αlarge␈αorganizations␈αor␈α
administrative
␈↓ ↓H␈↓computing␈α
and␈α
may␈α
be␈α
discounted␈α
in␈α
deciding␈α
what␈α
has␈α
to␈α
be␈α
done␈α
to␈α
serve␈α
a␈α∞large␈α
community
␈↓ ↓H␈↓and␈αhow␈αto␈αdo␈αit␈αwith␈αas␈αfew␈αpeople␈αas␈αpossible.␈α On␈αthe␈αother␈αhand,␈αthe␈αgeneral␈αviews␈αwere␈αthe
␈↓ ↓H␈↓basis for setting up LOTS and running it.

␈↓ ↓H␈↓        I␈αwas␈αdisappointed␈αthat␈αyou␈αdid␈αnot␈αconsider␈αLOTS␈αand␈αthe␈αAI␈αLab␈α
su≠ciently␈αsuccessful
␈↓ ↓H␈↓to␈αwarrant␈α
including␈αme␈αin␈α
your␈αstudy,␈αbut␈α
that␈αis␈αyour␈α
prerogative.␈α What␈αmade␈α
me␈αangry␈αwas␈α
to
␈↓ ↓H␈↓receive␈α
an␈α∞invitation␈α
over␈α∞your␈α
signature␈α∞to␈α
the␈α
hearing␈α∞that␈α
did␈α∞not␈α
include␈α∞me␈α
as␈α∞Director␈α
of
␈↓ ↓H␈↓LOTS␈α∪or␈α∪even␈α∪a␈α∩phone␈α∪call␈α∪telling␈α∪me␈α∪why␈α∩not.␈α∪ When␈α∪the␈α∪external␈α∪Advisory␈α∩Committee
␈↓ ↓H␈↓invitation␈α
also␈α
did␈α
not␈α
include␈α
me,␈α
I␈α
mistakenly␈α
concluded␈α
that␈α
there␈α
was␈α
a␈α
positive␈α
desire␈α
to␈α
avoid
␈↓ ↓H␈↓my point of view.

␈↓ ↓H␈↓        I␈α
have␈α
since␈α
learned␈α
that␈α
the␈α∞list␈α
for␈α
the␈α
Advisory␈α
Committee␈α
was␈α
made␈α∞by␈α
Feigenbaum's
␈↓ ↓H␈↓new␈α
secretary,␈αand␈α
the␈αCommittee␈α
made␈α
amends␈αby␈α
inviting␈αme␈α
to␈α
dinner.␈α This␈α
encouraged␈αme␈α
to
␈↓ ↓H␈↓call␈αJon␈αSandelin␈αand␈αtell␈αhim␈αthat␈αI␈αwould␈αspeak␈αfor␈αLOTS␈αunless␈αspeci≡cally␈αasked␈αnot␈αto,␈αand
␈↓ ↓H␈↓he apologized for what he said was his own mistake.

␈↓ ↓H␈↓        Ralph␈α
has␈α
done␈α
an␈α
excellent␈α
job␈α∞as␈α
Manager␈α
of␈α
LOTS,␈α
and␈α
when␈α
occasions␈α∞for␈α
publicity
␈↓ ↓H␈↓have␈αarisen,␈αI␈αhave␈αbeen␈αcontent␈αto␈αstay␈αat␈αthe␈αAI␈αLab␈αand␈αlet␈αhim␈αbe␈αinterviewed.␈α Nevertheless,
␈↓ ↓H␈↓I␈α
have␈α
ful≡lled␈αmy␈α
responsibilities␈α
as␈αDirector,␈α
and␈α
I␈αhave␈α
ideas␈α
and␈αproposals␈α
for␈α
the␈α
future␈αof
␈↓ ↓H␈↓LOTS that I would like to implement once we have solved our usage crunch.

␈↓ ↓H␈↓        Even␈αwith␈αthe␈αclari≡cation␈αthat␈αhas␈αoccurred,␈α
I␈αstill␈αam␈αnot␈αcon≡dent␈αthat␈αmy␈αviews␈α
on␈αthe
␈↓ ↓H␈↓future␈α∞of␈α∞computing␈α∞at␈α∞Stanford␈α∞in␈α∞general␈α∞or␈α∞even␈α∞at␈α∞LOTS␈α∞are␈α∞su≠ciently␈α∞wanted␈α∞so␈α∂that␈α∞it
␈↓ ↓H␈↓won't waste your time, the committee's time and mine to express them.


␈↓ ↓H␈↓Best regards,



␈↓ ↓H␈↓John McCarthy, Director of LOTS