Computerized Decision Making

Computerized decision making has taken off in recent years. Some have blamed the worldwide financial crisis on excessive reliance on these computerized decision making models. Lending officers who previously made individualized decisions about credit worthiness through personal judgment were replaced by computerized statistical models, resulting in mechanistic decision making. Large numbers of decisions were tied to a common set of assumptions, and when those assumptions proved to be wrong, the entire credit system fell apart and the economy faltered.

Use of computerized systems like Twitter and apps for information sharing may be leading to information overload. Eric Kessler from Pace University's Lubin School of Business notes, "What starts driving decisions is the urgent rather than the important." Researchers are finding that people who use too much information actually make worse decisions than people with less information, or they get so swamped in information that they are unable to reach a decision at all.

Computer decision models present certain advantages. Computers can amass and compile enormous amounts of data to spot patterns a human observer would never see. Computers are not prone to emotional decision making or the heuristics and biases we discussed in this chapter. Finally, computerized decision making systems are generally faster than human beings.

Computer decision making systems have certain faults that might severely constrain their usefulness; however, they are not capable of intuition or creative thought. As scholar Amar Bhidé notes, "An innovator cannot simply rely on historical patterns in placing bets on future opportunities." People are much more likely to spot opportunities that lie just beyond what the data can tell us directly. Also, groups of people working in collaboration can discuss and question assumptions and conclusions.

Computers cannot consider whether their programming makes sense or adapt automatically when values change. There are computerized aids for almost any decision to be made; each of the tools from simple lists to complex statistical models needs to be considered in light of the application. Managers who use computerized decision making need to take great care in determining exactly what type of output will truly meet their informational needs so they--not their machines--can make the best organizational decisions.

1 Give an example of a time where you used computerized decision making making sure to describe any advantages/benefits or disadvantages/challenges that you experienced.