84 1541-1672/08/$25.00 © 2008 IEEE IEEE INTELLIGENT SYSTEMS Published by the IEEE Computer Society Human-Centered Computing door. A number of social, economic, technological, and scientific trends have led to the emergence of communities of practice centered on the notion of the knowledge-based organization. However, the scientific foundation (knowl- edge elicitation methodology) and the commercial growth of knowledge management (KM) have largely developed in parallel—that is, independently. So, the creation of human-centered systems faces lingering challenges. For each challenge, we ask, is this a matter of building intelli- gent technologies or of using technology intelligently? Background As the expert-systems field morphed into what’s now broadly called intelligent systems, it became clear that we could preserve corporate knowledge by using the knowledge-elicitation methods that had been used to cre- ate knowledge bases and inference engines. 2 For all their limitations and brittleness, expert systems had pointed to the idea that organizations might create knowledge repositories. 3 We could then use the knowledge bases, including corporate “lessons learned,” in training and corporate problem solving. 4 Norman Kamilkow, editor in chief of Learning Officer Magazine, seems to agree. He said, “There is a growing role for a chief learning officer type within enterprise-level companies. … There is a need to have somebody focused on how to keep the skills of the corporation’s work force at a high level.” 5 In the KM process, company management establishes a program whereby experts who possess valuable knowl- edge collaborate with a knowledge engineer. Working to- gether, they elicit the expert’s wisdom for inclusion in the organization’s knowledge base. In extreme cases, such as when a senior expert with specialized knowledge is soon to retire, the organization might retain or bring back the individual as a consultant. 3 The KM literature on business management and the trade press on KM suggest that a wave of enthusiasm about KM hit in the 1990s but was followed by some dis- appointment. 1 The disappointment might have stemmed largely from limited KM software solutions, overzealous software sales personnel, or merely poor project imple- mentations. For example, of over 220 KM implementa- tions in 2000, at least half were “deeply suboptimized.” 6 Certainly, issues of lack of trust and perceived effective- ness were and are in play. 7 However, at least some disap- pointment could certainly be due to failure to properly embed these KM software systems in the human activities and work processes they were intended to support—that is, lack of human-centering. Whatever the reasons were, various trends and forces have encouraged a renewed in- terest in KM. Workforce issues Workforce mobility and its implications for the trans- fer of expertise have made KM a hot topic ever since this HCC department last discussed it. 1 For example, in even the most highly technical military jobs, the tradition of regular change of duty assignment requires considerable relearning. Just when a weather forecaster achieves jour- neyman-level skill at one locale, the Navy transfers him or her to some other climate. 8,9 Knowledge Management Revisited Robert R. Hoffman, Institute for Human and Machine Cognition David Ziebell, Electric Power Research Institute Stephen M. Fiore, Institute for Simulation and Training and University of Central Florida Irma Becerra-Fernandez, Florida International University A previous essay in this department 1 described how organizations are finding themselves in catch-up mode. They’re losing their ability to conduct business as their workforce ages and critical knowledge walks out the Editors: Robert R. Hoffman, Patrick J. Hayes, and Kenneth M. Ford Institute for Human and Machine Cognition rhoffman@ihmc.us