/LMAR=0/XLINE=3/FONT#0=BAXL30/FONT#1=BAXM30/FONT#2=BAXB30/FONT#3=SUB/FONT#4=SUP/FONT#5=BASL35/FONT#6=NGR25/FONT#7=MATH30/FONT#8=FIX25/FONT#9=GRK30



␈↓ ↓H␈↓α␈↓ ∧(COMMENTS ON DERTOUZOS'S PAPER

␈↓ ↓H␈↓␈↓ α_I␈αagree␈α
with␈αDertouzos's␈α
fundamental␈αcontention␈α
that␈αinformation␈α
processing␈αtechnology␈α
will
␈↓ ↓H␈↓enable␈α⊃us␈α⊃␈↓↓"to␈α⊃reverse␈α⊃some␈α⊃of␈α⊂the␈α⊃impersonal␈α⊃and␈α⊃dehumanizing␈α⊃consequences␈α⊃of␈α⊃the␈α⊂industrial
␈↓ ↓H␈↓↓revolution"␈↓.␈α -␈αat␈αleast␈αto␈αreverse␈αwhat␈αare␈αperceived␈αas␈αsuch␈αconsequences.␈α We␈αshall␈αindeed␈αhave
␈↓ ↓H␈↓individually␈α
designed␈αshoes␈α
and␈αclothes␈α
at␈αa␈α
price␈α
close␈αto␈α
that␈αof␈α
the␈αmass␈α
produced␈αproduct.␈α
 On
␈↓ ↓H␈↓the other hand I find the detailed picture rather implausible for reasons I can't always quite state.

␈↓ ↓H␈↓␈↓ α_My␈α→basic␈α→problem␈α→is␈α→that␈α~Dertouzos␈α→sees␈α→the␈α→various␈α→marvels␈α~as␈α→straightforward
␈↓ ↓H␈↓improvements␈αof␈αpresent␈αtrends.␈α Therefore,␈αhe␈α
tends␈αto␈αproject␈αpresent␈αideas␈αto␈α
the␈αachievement
␈↓ ↓H␈↓of␈α
these␈α
goals.␈α∞ I␈α
think␈α
we␈α∞are␈α
at␈α
least␈α
one␈α∞scientific␈α
revolution␈α
away␈α∞from␈α
the␈α
level␈α∞of␈α
artificial
␈↓ ↓H␈↓intelligence␈α⊂required␈α⊂to␈α⊃achieve␈α⊂many␈α⊂of␈α⊃the␈α⊂objectives␈α⊂present␈α⊃in␈α⊂this␈α⊂paper.␈α⊃ Therefore,␈α⊂the
␈↓ ↓H␈↓future will be more intellectually surprising than the paper suggests.

␈↓ ↓H␈↓␈↓ α_Here are some specific disagreements, mostly minor:

␈↓ ↓H␈↓␈↓ α_1.␈αLike␈αmost␈αwriters,␈αDertouzos␈αjumps␈αto␈αconclusions␈αabout␈αthe␈αrelation␈αbetween␈αtechnology
␈↓ ↓H␈↓and␈αemployment.␈α Looking␈αat␈αthe␈αpast,␈αAmerican␈αindustrial␈αproductivity␈αhas␈αgrown␈αby␈αmore␈αthan
␈↓ ↓H␈↓an␈α
order␈α
of␈α
magnitude␈α
since␈α
1900.␈α
 Unemployment␈α
levels␈α
are␈α
similar,␈α
and␈α
their␈α
is␈α
no␈αcorrelation
␈↓ ↓H␈↓between␈α↔specific␈α↔advances␈α↔in␈α↔technology␈α_and␈α↔the␈α↔level␈α↔of␈α↔unemployment.␈α_ The␈α↔economic
␈↓ ↓H␈↓servomechanisms␈α∀governing␈α∀employment␈α∀have␈α∃a␈α∀much␈α∀shorter␈α∀time-scale␈α∃than␈α∀technological
␈↓ ↓H␈↓advance.␈α∂ They␈α∂can␈α∂handle␈α∂much␈α∂faster␈α∂rates␈α⊂of␈α∂advance␈α∂than␈α∂are␈α∂likely␈α∂in␈α∂this␈α⊂country.␈α∂ For
␈↓ ↓H␈↓example,␈α∞they␈α∞can␈α∞handle␈α∞the␈α∞ten␈α∞percent␈α∞per␈α∞year␈α∞advance␈α∞that␈α∞Japan␈α∞experienced␈α∂during␈α∞the
␈↓ ↓H␈↓1960s.␈α The␈αerror␈αis␈αlike␈αsomeone␈αtrying␈αto␈αpredict␈αthe␈αeffect␈αof␈αa␈αchange␈αin␈αthe␈αgain␈αof␈αa␈αstage␈αof
␈↓ ↓H␈↓a␈α⊃feedback␈α⊃amplifier␈α⊃on␈α⊃the␈α⊃output.␈α⊃ There␈α∩is␈α⊃an␈α⊃effect,␈α⊃but␈α⊃it␈α⊃is␈α⊃completely␈α⊃masked␈α∩by␈α⊃the
␈↓ ↓H␈↓feedback loop.  Unfortunately, we don't yet know how to control the economic feedback loop.

␈↓ ↓H␈↓␈↓ α_2.␈α∩Count␈α∩me␈α∩with␈α∪those␈α∩who␈α∩believe␈α∩there␈α∪is␈α∩nothing␈α∩qualitatively␈α∩new␈α∪in␈α∩connecting
␈↓ ↓H␈↓computers␈α∂together.␈α∂ I␈α∂think␈α⊂one␈α∂can␈α∂prove␈α∂mathematically␈α⊂that␈α∂one␈α∂computer␈α∂can␈α⊂simulate␈α∂n-
␈↓ ↓H␈↓interacting computers at a speed of approximately 1/n times real time.

␈↓ ↓H␈↓␈↓ α_3.␈α∩While␈α⊃"arithmetic␈α∩atrophy"␈α∩of␈α⊃some␈α∩kind␈α∩is␈α⊃a␈α∩plausible␈α∩conjecture,␈α⊃I'll␈α∩bet␈α∩␈↓∧1␈↓50␈α⊃that
␈↓ ↓H␈↓Dertouzos can't cite a study actually observing it (written before December 1977).

␈↓ ↓H␈↓␈↓ α_4.␈α
One␈α
false␈α
note␈α
in␈α
the␈α
shoe␈α
scenario␈α
is␈α
the␈α
card.␈α
 The␈α
home␈α
terminal␈α
and␈α
personal␈α
data
␈↓ ↓H␈↓files␈α∞will␈α
almost␈α∞certainly␈α
precede␈α∞the␈α
shoe-maker,␈α∞so␈α
one's␈α∞measurements␈α
will␈α∞be␈α
stored␈α∞in␈α
one's
␈↓ ↓H␈↓home files and so will the results of new measurements.