Seeking the equivalent of the GUI for knowledge applications
Business is evolving inexorably toward the cyberorganization--a world of activities that integrates knowledge work and the networked computing environment and, in the process, changes the nature of work itself as well as the technological solutions applied to those activities.
Knowledge management is, in part, the inevitable response to that evolution. KM marks a shift from using computers solely as tools to the new role of the networked computer and software as pervasive organizational infrastructure. The most intriguing question in KM may well be how software vendors respond to that challenge during the coming years, because the change can best be likened to the transition from isolated character-based applications to graphical user interface (GUI) environments in the late 1980s and early 1990s. <" Images/towardacoalitionanearlymeet.gif"
The knowledge-based business environment
Although consultants are willing to talk endlessly about knowledge management as the business strategy of the next millennium, we may be spending too little time thinking about the full implications of the knowledge-based business environment and about the computer tools needed to address that environment.
Computers are replacing face-to-face exchanges, paper-based tools and publishing as the primary methods and media for creating, managing and transferring knowledge. The networked computing environment is the medium of our work--not just the tools of our work.
Knowledge-related tasks are driven to ever-lower levels of the organization. Managers often do not know how to improve the effectiveness of their knowledge workers because they do not understand the "stuff" of knowledge--and how technology can be applied to it.
Knowledge-related problems and activities characterize our work, and they change with amazing speed. Yet there is less time than ever to learn how to solve problems and develop expertise in those activities. It is often not possible for one person to find, absorb and test the information he or she needs to solve new problems effectively, and it makes even less economic sense to have many people duplicating such efforts.
We often cannot tell when a decision--based on the application of knowledge--is the right one. That is different from physical work or other activities in which there is feedback. A programmer receives meaningful, consistent feedback from many kinds of programming activities. But not the average knowledge worker.
Knowledge management is, in part, a recognition of the desperate need for a centripetal, integrative force in business that counteracts the forces of infoglut and technology. Technology itself is essential in solving those problems, but are technology vendors meeting the requirements of knowledge-based business?
A key to answering that question can be found in the history of the rise of GUI software to dominance over character-based applications.
Nothing was wrong inherently with the DOS version of Paradox (from Borland, Scotts Valley, CA), Lotus' (Cambridge, MA) Magellan, Symantec's (Cupertino, CA) GrandView outline processor or hundreds of other character-based applications. Those programs were not consigned to the status of pleasant memories because they lacked ease of use. They were more powerful and often easier to use than many of the Windows programs that supplanted them in the marketplace ... when they were used in isolation.
But in the business context of working with multiple applications, their lack of a common model for such basic operations as opening and closing files, selecting and copying text, printing, and exchanging data among applications made them burdensome.
Similarly, there's nothing inherently wrong with today's image management, document management and full-text retrieval applications ... in isolation from the knowledge management requirements of organizations as a whole. Yet they could still become nostalgia software like Paradox and Magellan.
The nature of work has expanded to include new practices and techniques for the development, processing and transfer of knowledge among a growing number of employees. And our time at work is spent not just processing information with a variety of tools, but understanding, defining and making judgments about knowledge-based work activities.
The proliferation of data processing, word processing and related requirements made the uniform graphical interface essential. The proliferation of knowledge-dependent activities and corresponding technologies imposes expanded or new requirements on the cyberorganization.
We must expand access to information, but we also must reduce the burdens of finding and using that information through widespread filtering, summarization and abstraction.
So much knowledge is available and needed that searching, managing documents, building knowledge and transferring knowledge value must be treated collaboratively rather than as isolated activities. Assessments of the value of information must also be shared.
The "context" of the user must be identified to provide the most appropriate knowledge to that user--just-in-time knowledge delivery. Conversely, lessons learned from the application of knowledge must be fed back into the knowledge base itself.
Publications and documents in the traditional sense are not what we want. We need dialog-based interaction with knowledge bases that parallel our person-to-person dialogs with subject matter experts. And we must become savvy about the use of knowledge.
Some characteristics of the evolving cyberorganization are already evident in content management and knowledge management technology.
Information avoidance. The extensive use of meta data in document management and other technologies is key to receiving what you want and avoiding what you don't need. Similarly, advanced full-text retrieval is the foundation for filtering and automatic summarization.
Relationships as bearers of knowledge. Hypertext links on the Web tell you that part of one document is connected by meaning to another. The simple relationships of meaning expressed by WWW links barely hint at the potential of computer-supported connections to convey--and preserve--valuable knowledge.
Collaborative construction of knowledge. In medicine, academia and commercial domains, technology is being applied to distributed, collaborative construction of knowledge. Traditional clinical research and individual research models can't stand up under the combined strains of infoglut, pace of change and demand for results.
New "publishing" model. Those same strains, added to the increasing complexity of our business and personal lives, make documents (even electronic documents) a bad fit for our knowledge transfer requirements. We need answers--not documents. Technologies like solution-centered support are beginning to solve the problems of providing only the information needed, when and where it is needed.
Vendors who are responding to those requirements are not just relabeling old information products to catch the wave of knowledge management. They're trying to adapt to a new business reality that they played major roles in creating, directly or indirectly.
What's next--from GUI to CIM?
Quarterdeck's (Marina del Rey, CA) Desqview was a wonderful environment for swapping DOS applications. Lightning-fast compared to the early versions of Windows, it was also a delight to use, until you wanted the applications to really work with each other, or work like each other.
There's much to be said for the ease of use of graphical user interfaces, but the core benefit of the Mac, Windows, Motif and other GUIs is providing a "usability infrastructure." Common activities that were treated differently in character-based applications (even figuring out how to exit some programs was a problem) become second nature for users of Windows or Mac applications. The GUIs allow users to focus more on what they want to do than on the operation of tools.
The common features of using computer applications as tools for specific tasks were identified, distilled and placed at a different level--at the level of an overall environment, not at the level of applications.
The problem was solved at the meta level. Ed Swanstrom, founder of the Knowledge Management Consortium and chairman of Agilis (Gaithersburg, MD), points to the importance of understanding and solving problems by stepping up to a higher level: "Each meta level has, as the subject of discussion or study, a level below it," said Swanstrom.
Stepping up to a higher level will undoubtedly occur within computer technology for knowledge management. Perhaps we'll need a "concept information manager"--a service layer and infrastructure for the pervasive tasks of managing knowledge. Some of those services are already suggested by computer-supported relationships (links) as bearers of meaning, collaborative development of knowledge resources and dialog-like queries of knowledge bases.
Unfortunately, most of those are embedded in computer resources that are delivered to us as finished products--hypertext links, live tables of contents in documents, knowledge bases that can be queried as if they were subject matter experts.
We don't have access on the desktop to the tools needed to do similar things. For example, clipping text with embedded links from Web pages helps us capture a computer-supported connection. But what if the relationship is expressed in some other way--for example, by a link from a frame to an HTML page? What if the link is more than a simple jump to another page, expressing instead, for example, a hierarchical or a cause/effect relationship between one content element and another?
Challenges and opportunities for vendors
Is this yet another opportunity for Microsoft (Redmond, WA), Sun (Mountain View, CA) and Apple (Cupertino, CA) to build application features into their operating systems? If they're smart, they will. But they will do so selectively, after they learn about the impact of knowledge management on computing--a set of changes in business that OS vendors are heavily responsible for creating.
But we still have relational database products, including Paradox itself in its Windows incarnation, and utilities with some of the functions of Magellan.
The developer community, driven by the knowledge management needs of its customers, must identify the essential characteristics and benefits of current KM technologies, including document management and full-text retrieval, from the perspective of the needs in the business community. The vendor community should remember that developers of GUI software have benefited from using a common service layer for file access and other functions in the form of greater functionality, elimination of most concerns with peripherals, lower cost of code maintenance and reduced support requirements.
Similarly, if Microsoft, Sun and others find a way to provide a usability infrastructure for KM applications, they should do so. Application developers will benefit, too. OS vendors would be well-advised to create partnerships with and learn from application developers, because the need for intelligent handling of knowledge is so pervasive and diverse that they cannot own it. The business problems associated with knowledge management are far more complex than can be solved with generic OS replacements for highly sophisticated discrete applications.
Identifying customer needs
Vendors also need to take another close, critical look at what their customers are doing in their knowledge work. Vendors can no longer view their applications in isolation from other aspects of the knowledge-based cyberorganization.
Application vendors are forgetting the special competencies they developed through identifying, responding to and satisfying customer requirements. Those special competencies distinguish vendors who enjoy enduring success from those who fail. Slight technical edges are not a major factor.
Vendors of document management, workflow, imaging and retrieval software can use their historical competence in meeting customer business requirements to help them understand the new roles KM applications play in the evolving cyberorganization. And vendors need to help customers identify how their technology meets new business requirements within a partnership of technologies that must be integrated at a higher level.