THE RISE OF THE EXPERT COMPANY How...

THE RISE OF THE EXPERT COMPANY How Visionary Companies Are Using Computers to Make Huge Profits by Edward Feigenbaum, Pamela McCorduck and H. Penny Nii (Times Books: $18.95) The office secretary trying to coax a screeching dot matrix printer out of crumpling yet another sheet of paper might wonder whether computers are indeed bringing a revolution of efficiency to the workplace, as these authors claim. But while high-tech progress has sometimes been slow to trickle down to the microcomputer (PC) level, this book leaves little doubt that advances such as artificial intelligence (AI) are dramatically changing the way some major American corporations are conceiving of and carrying out work.

At American Express, for example, AI computers can analyze a customer’s spending habits over the past week and make a complex decision within a second about whether to approve an individual purchase. The most striking changes chronicled here, however, are future applications of AI. Libraries, for instance, will likely become “active, intelligent knowledge servers,” the authors write, enabling users to call up or compare books by ideas rather than titles.

World competition for the AI market is understandably intense, for as MIT economist Lester Thurow has pointed out, “Standards of living rise not because people work harder but because they work smarter.” Bolstered by great technical strength and 20 years of Defense Department funding for AI research, Americans lead the race, the authors report, followed by the Japanese (whose government is coordinating the ambitious building of a “knowledge industry”) and the European Economic Community (with smaller budgets but excellent leadership).

The authors’ perspective is sometimes obscured by their enthusiasm. It is misleading, for example, to claim that “The hallmark of knowledge processing is reasoning by computer,” when the systems they profile in these pages cannot “reason” as we know it, using abstraction and intuition. And while the authors are generally diligent about reporting the politics behind the implementation of AI, they devote only one vague paragraph to the important, troubling question of what will happen to workers whose jobs are replaced by AI computers. Overall, though, this is a fine book, the first to show lay readers the practical applications of AI in the 1980s.