on Herbert Simon, artificial intelligence pioneer, who died on February 9th, aged 84:
„The story goes that one day
Herbert
Simon announced to a group of his students that he and some of his
colleagues
had invented a "thinking machine". He said it was equal, and
perhaps
superior, to the human brain.
The idea of such a marvellous
rnachine
had been around for centuries. René Descartes (1596-1650),
whose philosophy was bound up with rnathematics, wondered whether a
machine
could be made to think; but generally the notion was considered as
unlikely
as turning iron into gold, another fantasy that occupied great
minds.
Now here was Mr Simon disturbing his class on a sleepy afternoon in
Chicago
in 1957 with his improbable claim for computers.
At that time the computer was
chiefly
famous for having been invented by the British to decode enemy messages
in the second world war. High-tech in the world at large was not
much more than the electric typewriter. But for several years Mr Simon
had been examining whether a computer could match the process of human
thought. As a test, he showed that a computer could quickly
provide
a proof of theorems in "Principia Mathematica", a key 20th-century work
on logical theory. Bertrand Russell, one of the book's authors,
said
a trifle testily that he wished he had known what "can now be done by
machinery"
before he had „wasted ten years doing it by hand."
But can artificial intelligence,
AI as insiders call it, really be dignified with, well, "intelligence"?
Is this what Descartes had in mind when he wrote, "I think, therefore I
am"? A machine will deliver a bar of chocolate in return for a
coin,
provided it is working, but not even a child believes it is a thinking
instrument. A computer is a machine, however intricate. Is it no
more than a numnber cruncher? For more than 40 years Mr Simon was
regularly interrogated about bis claim that there was little difference
between a suitably programmed computer and the human brain's use of
neurons.
He usually answered questions with good grace, but he refused to give
ground.
Obvious once you know
In 1978, Herbert Simon was awarded the
Nobel prize for economics. What for many people would be regarded
as the culmination of a life's work, Mr Simon took almost casually, a
diversion.
The Swedish judges at the presentation ceremony were a touch hurt to
hear
that artificial intelligence had been his central interest, rather than
economics, although of course he was interested in that discipline too.
But to those who knew him such versatility was no great surprise. He
dabbled
in many things, usually with great accomplishment. What he called
"social science" took a hold on him, but he could probably have made a
career as a pianist or a painter.
His parents were German
immigrants who, like many before them, had settled in Milwaukee.
While still at university he had a part-time job with Milwaukee's local
authority and became interested in how the administration made budget
decisions,
or choices as Mr Simon preferred to call them. Years later
"Administrative
Behaviour" was the subject of his doctorate, and later still the
dissertation
was turned into a book of the same name, probably the best known of Mr
Simon's 20 books. It was his ideas on decision-making, especially in
business,
that caught the eye of the Nobel judges.
Like many economic
theories, Mr Simon's seems obvious once you know it. In taking a
decision,
he said, no business could process satisfactorily all the "zillion
things"
affecting the marketing of a product, in the hope that the right answer
for maximising profit would pop out at the end.
That was classical
cconomic theory, he said, but it was "a ridiculous view of what goes
on".
Rather, a business tried to make a decision that was "good enough". He
called his theory "bounded rationality" and invented a name to describe
it: "satisficing", a composition of the words satisfy and
suffice.
Not all economists agreed with Mr Simon. "But they are mistaken,"
he said.
His views on economics tied in with
his ideas on artificial intelligence. Even a computer displayed
its
intelligence by making choices, he said. Like a human, a chess
computer
would analyse the consequences of a move, but it would do better than
even
a grandmaster, who would be unlikely to see beyond eight moves
ahead.
But what about insight? Or indeed wisdom and creativity? Mr Simon
tended to be dismissive of such vague human terms. His computers
had created drawings, which he was happy to display in his office, and
music, which musicians said had aesthetic interest. They had made
choices, as a human artist or musician would.
For many people, artificial
intelligence
suggests Hal, the worryingly clever human-like computer which rebelled
in the film „2001“. Although Mr Simon sometimes seemed to suggest
that a Hal was just around the comer, he was not going to be drawn into
comparisons with science fiction. His strictly scientific aims,
he
said, were limited to using cornputers to understand how humans think,
and as an aid to human thinking. What about the soul? No one, he
said, would tell him what the soul was. When someone did, he said
thoughtfully, he would program one.“
© The Economist, London