291
a form, that when it shows elements
of other categories of learning
we may reasonably suspect
these of contamination with
a trial-and-error component.”
Turning point
Before developing MENACE,
Donald Michie had pursued a
distinguished research career in
biology, surgery, genetics, and
embryology. After MENACE, he
moved into the fast-developing
area of artificial intelligence (AI).
He developed his machine learning
ideas into “industrial-strength
tools” applied in hundreds of
situations, including assembly
lines, factory production, and steel
mills. As computers spread, his
artificial intelligence work was
used to design computer programs
and control structures that could
learn in ways perhaps not even
guessed at by their human
originators. Michie demonstrated
that careful application of
human intelligence empowered
machines to make themselves
smarter. Recent developments
in AI use similar principles to
develop networks that mirror the
neural networks of animals’ brains.
Michie also conceived the
notion of memoization, in which
the result of each set of inputs in a
machine or computer was stored as
a reminder or “memo.” If the same
set of inputs recurred, the device
would at once activate the memo
and recall the answer, rather than
recalculating afresh, thereby saving
time and resources. He contributed
the memoization technique to
computer programming languages
such as POP-2 and LISP. ■FUNDAMENTAL BUILDING BLOCKS
New computer technology has led to
a rapid development in AI, and in 1997,
the chess machine Deep Blue defeated
world champion Garry Kasparov.
The computer learned strategy by
analyzing thousands of past games.He had this concept that
he wanted to try out that he
thought might possibly solve
computer chess...It was the
idea of reaching a steady state.
Kathleen Spracklen