Moshe Y. Vardi
Chess fans remember many dramatic chess
matches in the 20th century. I recall being
transfixed by the 1972 interminable match
between challenger Bobby Fischer and
Past and future
defending champion Boris Spassky for
the World Chess Championship. The
most dramatic chess match of the 20th
century was, in my opinion, the May
1997 rematch between the IBM super-computer Deep Blue and world champion Garry Kasparov, which Deep Blue
won 3½– 2½.
I was invited by IBM to attend the rematch. I flew to New York City to watch
the first game, which Kasparov won. I
was swayed by Kasparov’s confidence
and decided to go back to Houston,
missing the dramatic second game, in
which Kasparaov lost—both the game
and his confidence.
While this victory of machine over
man was considered by many a triumph for artificial intelligence (AI),
John McCarthy (Sept. 4, 1927−Oct.
24, 2011), who not only was one of the
founding pioneers of AI but also coined
the very name of the field, was rather
dismissive of this accomplishment.
“The fixation of most computer chess
work on success in tournament play
has come at scientific cost,” he argued.
McCarthy was disappointed by the fact
that the key to Deep Blue’s success was
its sheer compute power rather than a
deep understanding, exhibited by expert chess players, of the game itself.
AI’s next major milestone occurred
last February with IBM’s Watson pro-
gram winning a “Jeopardy!” match
against Brad Rutter, the biggest all-time
money winner, and Ken Jennings, the
record holder for the longest champion-
ship streak. This achievement was also
dismissed by some. “Watson doesn’t
know it won on “Jeopardy!”,” argued the
philosopher John Searle, asserting that
“IBM invented an ingenious program,
not a computer that can think.”
In fact, AI has been controversial
from its early days. Many of its early pio-
neers overpromised. “Machines will be
capable, within 20 years, of doing any
work a man can do,” wrote Herbert Si-
mon in 1965. At the same time, AI’s ac-
complishments tended to be underap-
preciated. “As soon as it works, no one
calls it AI anymore,” complained McCar-
thy. Yet it is recent worries about AI that
indicate, I believe, how far AI as come.
Moshe Y. Vardi, EDITOR-IN-CHIEf