Decisions, Decisions

W. Brian Arthur

Citibank Professor
Santa Fe Institute
Santa Fe, New Mexico

I received my PhD in operations research, which is a highly scientific, mathematical way of strategizing and of making decisions. I once thought that I could make any decision, whether professional or personal, by using decision trees, game theory, and optimization. Over time, I've changed my mind. For the day-to-day work of running a business - scheduling a fleet of oil tankers, choosing where to open a new factory - scientific decision theory works pretty well. But for just about every other kind of decision, it doesn't work at all.

For the big decisions in life, you need to reach a deeper region of consciousness. Making decisions then becomes not so much about "deciding" as about letting an inner wisdom emerge. We've been bamboozled into believing that cognition is rational - that our mind is a gigantic computer, or a blackboard on which we can reach a decision by calculating pluses and minuses. Recent research on cognition shows that our minds rarely make strictly logical deductions. Instead, we rely on patterns - and on feelings associated with those patterns.

So for those big decisions - Should I marry this person? Should I follow that career? Should I sell my company? When should we go public? - let patterns develop in your mind. Let clues and evidence emerge from your environment. This approach to decision making requires time, patience, and another key ingredient: courage. It takes courage to listen to your inner wisdom. But once you hear that wisdom, making a decision becomes fairly easy.

W. Brian Arthur's work on the economics of increasing returns and their role in technology-based industries has won him international recognition in the scientific and business communities. He is a pioneer of the new science of complexity and the author of The Economy as an Evolving Complex System II (Addison-Wesley, 1997).

Chung-Jen Tan

Senior Manager, Application Systems
IBM's Thomas J. Watson Research Center
Yorktown Heights, New York

I've learned a lot about decision making by watching how a machine plays chess. In 1996, Deep Blue played Garry Kasparov, the world chess champion. Deep Blue won the first game but lost the match. That experience taught us what we needed to do to be ready for the rematch in 1997.

A human uses a combination of knowledge, strategy, and intuition to make chess decisions. A machine relies on brute computational power and on an ability to examine a tremendous amount of data. On average, the human mind can manage three or four positions per second. Deep Blue can evaluate 200 million alternatives per second. So the machine has a huge advantage. But Deep Blue also had weaknesses, and they centered on its inability to be flexible. Our challenge was to equip Deep Blue to make decisions like a human grandmaster.

We built in several programming tools that allowed us to adjust some of the decision parameters dynamically and to change strategies in between games. Joel Benjamin, the U.S. chess champion, played chess with Deep Blue for nine months - testing and tunneling the system - and we programmed some of his knowledge into the computer.

An interesting thing happened when Deep Blue met Kasparov for the rematch. While we had refined Deep Blue so that it could make decisions and play more like a human, Kasparov had refined his strategy so that he could play better against standard computer-chess programs. In game six, Deep Blue surprised Kasparov by sacrificing a knight to gain strategic advantage. Kasparov, who hadn't planned for such a decision, realized how inflexible his own strategy had become. We showed that a machine can reach a level of play comparable with that of a grandmaster.

Chung-Jen Tan manages IBM's Deep Blue computer-chess project. Since 1984, Tan has worked on the architecture development and machine design of highly parallel, scalable systems for solving large, complex problems.

Deborah Triant

CEO and President
Check Point Software Technologies Inc.
Redwood City, California

Making and implementing decisions boils down to a key ingredient: listening.

I often wonder why schools emphasize debating. Why not have listening classes as well? Debating is easy; listening with an open mind is not. The worst thing that you as a leader can do in the decision-making process is to voice your opinion before anyone else can. No matter how open and honest your people are, stating your opinion first will short-change the discussion process and taint what you hear later. I've learned this the hard way.