What to Read Next
Already a member?Sign in
For decades, futurists have anticipated the day when computers would relieve managers and professionals of the need to make certain types of decisions.1 Computer programs would analyze data and make sound judgments on such matters as how to configure a complex computer, how to diagnose and treat a patient’s illness or how to know when to stir a big vat of soup with little or no human help. But automated decision making has been slow to materialize. Many early artificial intelligence applications were just solutions looking for problems, contributing little to improved organizational performance.2 In medicine, for example, doctors showed little interest in having machines diagnose their patients’ diseases. In the business sector, even when expert systems were directed at real issues, extracting the right kind of specialized knowledge from seasoned decision makers and maintaining it over time proved to be more difficult than anticipated.
Even though the need for automated decision systems was recognized, full-blown decision-making systems were seen as impractical for use in business. So, during the 1970s, managers began to address this need by employing intelligence augmentation tools that provided managers and analysts with “decision support.”3 The idea was for the support system to help managers report, analyze and interpret data as opposed to actually making the business decisions. Although some decision support tools offered the potential for sophisticated statistical insight into business problems, they generally required skilled users to direct their use. The tools were usually not integrated with business applications. As a result, managers used them to help make decisions and then, if computers could help, used separate applications to carry out the decisions. For these and other reasons, such tools didn’t catch on — not nearly to the extent that more transactional software applications, such as enterprise resource-planning systems, did.
The reluctance on the part of executives to embrace decision-support tools during the 1970s and 1980s was not surprising.
Read the Full ArticleAlready a subscriber? Sign in
1. For example, scientist and fiction writer I. Asimov’s “I, Robot” (New York: Gnome Press, 1950) identified the Three Laws of Robotics, an early set of rules to guide automated decision making. R. Kurzweil’s “Age of Intelligent Machines” (Cambridge, Massachusetts: MIT Press, 1990) contains an overview of the history of artificial intelligence.
2. T.G. Gill, “Early Expert Systems: Where Are They Now?” MIS Quarterly 19, no. 1 (March 1995): 51–81.
3. Decision-support systems were defined in A.G. Gorry and M.S. Scott Morton, “A Framework for Management Information Systems,” Sloan Management Review 13, no. 1 (fall 1971): 55–70.
4. For a description of business-rules technology, see B. Von Halle, “Business Rules Applied: Building Better Systems Using the Business Rule Approach” (New York: John Wiley & Sons, 2001).
5. For an overview of yield-management applications in the transportation industry, see A. Ingold, U. McMahon-Beattie and I. Yeoman, “Yield Management” (New York: Continuum, 2001).
6. For more on automated decision making in the consumer credit industry, see T.H. Davenport and J.G. Harris, “Automated Decision Making in Consumer Lending,” research note, Accenture Institute for High Performance Business, New York, June 2004, www.accenture.com.
7. For more on automated decision making at DeepGreen Financial, see J.G. Harris and J.D. Brooks, “In the Mortgage Industry, IT Matters,” Mortgage Banking Magazine, Dec. 4, 2004, 62.
8. For more on automated decision making in the insurance industry, see T.H. Davenport and J.G. Harris, “Lessons For Successful Automated Decision Making From The Insurance Industry,” research note, Accenture Institute for High Performance Business, New York, November 2004, www.accenture.com.
9. “Pickberry Vineyard: Accenture Prototype Helps Improve Crop Management,” 2004, www.accenture.com.
10. T.H. Davenport and J. Glaser, “Just-in-Time Delivery Comes to Knowledge Management,” Harvard Business Review 80, no. 7 (July 2002): 107–111. See also D.W. Bates et al., “Effect of Computerized Physician Order Entry and a Team Intervention on Prevention of Serious Medication Errors,” Journal of the American Medical Association 280 (Oct. 21, 1998): 1311–1316.
11. In a previous study of how organizations build analytical capability, we found that quantitatively oriented experts were almost always present in organizations with high degrees of analytical activity. See T.H. Davenport, J.G. Harris, D.W. DeLong and A. Jacobson, “Data to Knowledge to Results: Building an Analytic Capability,” California Management Review 43 (winter 2001): 2, 117–138.
12. For more on how losing human expertise in a technologically intensive business can undermine organizational performance, see D.W. DeLong, “Lost Knowledge: Confronting the Threat of an Aging Workforce” (New York: Oxford University Press, 2004).