Computers are marvelous devices. With the right commands, they can mimic almost anything — from a musical keyboard to the chess playing abilities of a grand master
As computer scientists continue to explore the functional limits of this device, some computer fans not so subtly hint that this device will replace man in the intelligence hierarchy. As you watch this marvelous control undergo its next metamorphosis, (see “Fast Forward to the Future of Controls,” page 23), you may begin to agree. But, even though scientists work hard to overcome them, the computer — be it a microprocessor, PLC, or Cray — has limits.
Two of its limits were demonstrated in a huge battle for the title of intelligence supremacy waged in the middle of February 1996. The combatants were a chess grand master and the ultimate chessplaying computer, IBM’s Deep Blue.
The real battle, however, was man vs. man; between scientists’ efforts at mimicking human intelligence and an intelligent human.
The outcome was obvious before the first move. A computer can solve floating point calculations in thousandths of a second for smooth motion. It can cause various assembly tasks to happen in the proper sequence. It can gather all sorts of sensor and motion data, analyze it, and tell you where potential breakdowns may occur, where inefficiencies exist in a process, and how productive a factory was in a given time period.
But a computer cannot think like a human. And it may never be able to.
The field of computing has made great strides in developing machines that mimic some forms of human thinking. Chip developers and programmers have increased the speed at which a computer can carry out such thinking. (Deep Blue can review over 100 million chess positions a second). Scientists have even created computers that can deal with intangibles — fuzzy logic.
But the challenges of adaptive, creative, sudden-insightful thought have not been conquered by science. Deep Blue proved there’s still a ways to go in that exploration.
This battle showed two areas of computer vulnerability; human error, and having a strength turned to a weakness.
Regarding human error, the computer is limited to the degree its programmers are free of error.
• Deep Blue could not execute certain moves, because it couldn’t find the files containing that information. The files had not been loaded by the human programmers. • Humans did not properly execute chess moves Deep Blue calculated. • Software created to recognize certain strategies was never loaded, because the humans never expected to actually confront those situations.
Until man is incapable of error, a computer can never be error-free, for a computer is a mimicry of the minds that created it, subject to the same flaws and errors of its creators.
But even if the programmers had not made these mistakes, the outcome was still obvious — Deep Blue would still have lost. Mr. Kasparov understood how to turn a strength into a weakness. Said Mr. Kasparov in a TIME essay (3/25/1996), “At one point, I changed slightly the order of a well-known opening sequence. Because (the computer) was unable to compare this new position meaningfully with similar ones in its database, it had to start calculating away and was unable to find a good plan.” (Despite its speed, a human can still get a computer to calculate endlessly and fruitlessly).
Mr. Kasparov also stated that he avoided giving the computer any concrete goal to calculate toward; if it can’t find a way to win (chess pieces), attack the king, or fulfill one of its other programmed priorities, it drifts planlessly into trouble. He could figure out its priorities and adjust his play. Deep Blue couldn’t do the same to him. Despite human error, adaptive thought is still part of human, not silicon, intelligence.
As the computer continues to evolve to more nearly meet our needs for precise motion control, never forget that it has limits — it does exactly what you program it to do. If you don’t tell it to do something, it can’t “create” your intentions. Man still occupies the top spot on the intelligence hierarchy and will continue to until scientists solve some fundamentally human problems, like what is consciousness. Until then, enjoy the wonders of this device, but be aware it can’t do your thinking for you.
Leslie Langnau, Associate Editor