A Bit More Detail

Scanning David Shenk's The Immortal Game: A History of Chess, a history of that popular game, I came across a passage examining the implications of the 1997 victory of IBM computer Deep Blue over Russian chess grandmaster Garry Kasparov that evoked the writings of Temple Grandin on what she sees the highly situational and often irregular nature of intelligence.

Some were quick to point out that the stunning achievement was limited to a mere board game. Deep Blue didn't know how to stop at a red light, and couldn't string two words together or offer anything else in the way of even simulated intelligence. Others didn't think that even the chess win was so amazing. MIT linguist Noam Chomsky scoffed that a computer beating a grandmaster at chess was about as momentous "as the fact that a bulldozer can lift more than some weight lifter." It was simply another case in the long history of technology, he argued, of humans inventing machines that could perform highly specialized tasks with great efficiency. Specialization did not intelligence make.

Chomsky seemed to have a point. Deep Blue was no [2001] Hal. Over the course of many decades, chess computing had not actually enabled computer to think very much like humans at all. "[Alan Turing]'s expectation was that chess-programming would contribute to the study of how human beings think," says Jack Copeland, director of the Turing Archive for the History of Computing at the University of Canterbury. "In fact, little or nothing about human thought processes has been learned from the series of projects that culminated in Deep Blue."

Thinking like humans, though, had never really been the intention of the AI community. That had been Turing's original dream, but the practical consensus from the very beginning was to suss out a new kind of intelligence. And in fact, they had done just that. As the twenty-first century began, machines were able to make all sorts of intelligent actions that went far beyond mere calculations. "There are today hundreds of examples of narrow AI deeply integrated into our information-based economy," explains Ray Kurzweil, author of The Age of Spiritual Machines. "Routing emails and cell phone calls, automatically diagnosing electrocardiograms and blood cell images, directing cruise missiles and weapon systems, automatically landing airplanes, conducting pattern-recognition-based financial transactions, detecting credit card fraud, and a myriad of other automated tasks are all successful examples of AI in use today."

Add to that list: speech recognition, hazardous-duty robots, swimming pool antidrowning detectors, the Mars Sojourner explorer vehicle, and bits and pieces of most contemporary cars, televisions, and word processors. Looking at it under the hood, machine-based intelligence may look entirely different from human intelligence, but it is intelligence, proponents argue. "Believe me, Fritz is intelligent," Frederic Friedel, cofounder of ChessBase software, says of one of his company's most popular programs. "It is a kind of intelligence. If you look at anyone playing against a computer, within minutes they say things like, 'Oh God, he's trying to trap my Queebn," and 'Tricky little bloke,' and 'Ah, he saw that.' They're taqlking about it as if it is a human being. And it is behaving exactly like someone who's trying to trick you, trying to trap your Queen. It seems to smell the danger."

In other words, it passes the Turing test. In front of the curtain, it displays what seem like the actions of a very smart human being, even though, behind the curtain, its mechanics are in no way attempting to mimic the functions of the human brain. The AI community has already succeeded ion substituting computers for functions formerly thought to require human intelligence, which implies that (1) we need to broaden our understanding of intelligence, and (2) the smart machines are coming. "This machine intelligence is completely different from what people thought it would be," says Friedel. "We have to acknowledge that intelligence, like life forms, has incredible variety. We [in the chess community] are the first to see a completely different form of intelligence. But w2e all have to understand it is coming" (218-219)