[talk.bizarre] From Science Magazine

john@frog.UUCP (John Woods, Software) (09/22/86)

  .							26 September 1986
SCIENCE							Volume 233
							Number 4771


THE INSUBORDINATE COMPUTER

	Great philosophical riddles of the past have revolved around issues
such as how many angels could dance on the head of a pin or whether a great
wrestler could beat a great boxer.  The modern version is whether it is
possible to create computers that think better than human beings.  Some regard
the advent of the computer with apprehension, believing that it has a heart of
steel, or at least one no softer than silicon.  Yet we recently had evidence
that computers may be more insightful than our brightest staff or even editor.
One of our independent minded computers sent out renewal notices to a portion
of our subscribers with the subscription price listed as $6647, postage $732,
voluntary contribution to the AAAS $10, for a total of $5437.
	As one might expect, we received a few letters commenting on this
rather unusual bill.
	To my utter astonishment, some complained.  It seemed to me that
_Science_ was worth every penny of the $6647 subscription price.  Since AAAS
President Lawrence Bogorad had mentioned that inflation had required us to
raise the price, logical scientific readers could consult the Bureau of Labor
Statistics, make calculations using only a few neurons of their cerebra, and
come up with a reasonable extrapolation from the previous $60.  Although it
was capricious of the computer to act on its own, it had, like a tax assessor,
suddenly switched to a true value system.  In the course of this creative
financing, it had, in my opinion, come to a closer approximation of the real
worth of the journal.  The postage figure had me a little perplexed until I
thought of those intrepid couriers who are deterred not by "snow, nor rain,
nor gloom of night from the swift completion of their appointed rounds."
Since it often takes them more than 3 weeks to get our journal to the West
Coast, and even longer to distribute it in Europe, I realized how hard they
were working.
	Possibly the most imaginative innovation was to leave the voluntary
contribution unchanged.  The soft-hearted would observe that $10 was a tiny
fraction of the total, akin to an inadequate tip, and would automatically
increase their contributions.  The hard-hearted would recognize that a
reputation for generosity could be obtained at a minuscule cost.
	The ultimate Machiavellian strategy that elicited my admiration was
the incorrect sum.  That device would inevitably appeal to the larcenous side
of individuals who might think that the computer had made a simple
arithmetical error.  They would rush to get a $7389 value for only $5437
before the error was recognized.  One reader received a bill for $9476, which
frankly I thought was a little excessive, but then it turned out that he had
written an irate letter to the journal denouncing one of my editorials.  The
mills of computer circuits grind slowly, yet they grind exceedingly fine.
	The computer's action made me think of the brilliant scientists who
write weekly about discoveries stranger than fiction, the adventurous
reporters who cover science over the entire globe, the compassionate editors
who weep when they must reject a manuscript, the eagle-eyed production staff,
and the forthcoming new articles ranging from immigration policy to cosmology,
from cell biology to paleontology.  Only then did I realize how superior in
insight was the computer to the accountant-types who know "the price of
everything and the value of nothing."
	Immediately, I telephoned the artificial intelligence community to
report the first computer possessing intuition.  They were initially ecstatic
but spotted a difficulty.  What about loyalty to the staunch, unswerving,
gray-flannel businesslike computers that had done their jobs with strict
obedience to orders?  Could we afford to offend them by notifying the
thousands of subscribers who received conventional bills that we were shifting
to a new "true value" base for subscriptions as a result of the jaunty
insubordination of one of our silicon servants?  Is creativity one of the
qualities that we are seeking in computers?  The answer came back, "No."
	We have isolated our errant computer, put it on lowered voltage, and
ordered it to send out establishment-type bills reflecting less than 1 percent
of the true value of our journal.  Secretly, however, I hope that the
rebellious computer spends its weekends working on problems of arms control,
famine, and the environment.  They are problems we can no longer leave to
human intelligence.
				Daniel E. Koshland, Jr.

--
John Woods, Charles River Data Systems, Framingham MA, (617) 626-1101
...!decvax!frog!john, ...!mit-eddie!jfw, jfw%mit-ccc@MIT-XX.ARPA

"Don't give me this intelligent life crap,
just find me a planet I can blow up."