Question? Leave a message!

How Artificial Intelligence is helpful in Business Analysis

artificial intelligence in business decision making and artificial intelligence in business information systems | pdf free download
JohnTheron Profile Pic
Published Date:09-07-2017
Website URL
Knowledge Engineering 2017 The Uses of Artificial Intelligence in Business By Thomas B. Cross © 1 Chapter 1 Trends in Knowledge Engineering © 15 KNOWLEDGE ENGINEERING Introduction Knowledge engineering (KE) is the application of machine systems to problems of human endeavor. The purpose is not necessarily to develop systems that replace humans, but to allow the use of systems that increase human effectiveness and efficiency. The goal is to encourage humans to do what they do best, whatever this might be, at a time deemed appropriate, and to allow machines to assume the functions best suited to them, such as power saws, mechanical arms, and payroll computers do. KE has many major functions that are discussed here. Some of the fundamental KE issues addressed are: • Ergonomics: Human-machine interaction • Problem-solving and decision-making • Thinking • Communication: People-people, people-machine, and machine-machine • Dark-side problems (fear, conflict, privacy, etc.) • Intelligence • The future These issues are woven throughout the book, supplying the topical glue between distinct discussion areas. Of course, there are other areas related to these issues that are not covered here. It is the purpose of this book to focus on the concepts and issues of KE that will impact on business management strategies, productivity, and the key element of any business—its people. Much like auto mechanics determined to increase engine speed, those of us engaged in research in these areas often fail to present adequate and informative reasons for such enthusiastic pursuit. There are many humanistic and bottom-line reasons for explaining KE and its potential benefits to humankind. Experts, such as Herbert A. Simon, suggest that automation—the augmentation of manpower by machine power—is not the only way in which fundamental research in heuristic problem-solving is likely to contribute to productivity in our society. The most important productive resource in our economy and, very likely, the most important resource for generations to come, is brainpower. We are now learning a great deal about how this brainpower operates—about the processes of human thinking. This subtle, yet key issue is fundamental to any research in KE. It is the pursuit of humanistic uses of technology that offers the greatest challenge and hope for improving productivity and performance, while also improving the quality of life. © 16 TRENDS IN KNOWLEDGE ENGINEERING There are no right or wrong answers, only new problems to be solved. In the past, machines held forth the ability to free people from mundane, tedious, and repetitious tasks and chores. KE promises to facilitate human communication and interaction in ways unavailable or impossible without these systems. Moreover, as these systems become more "intelligent" and more competent, they should gradu- ally acquire the ability to make decisions on their own. The application of self-directed thinking machines offers mind-boggling benefits to education, quality control, and field engineering, to name only a few applications. The development of KE systems that can go places too toxic or otherwise unfit for humans is reason enough to pursue the development of such systems. Knowledge Engineering Limitations and Possibilities We now stand at the foot of the KE mountain, trying to find the most efficient path to the top. Today, the limitations of KE are associated with these general areas: • Computing capability • Problem definitions • Machine languages • Human knowledge ability Of these four areas, the basic horsepower problems of computing "engines" is the easiest to tackle. Computer processing capability has crossed the billion- instructions-per-second barrier. Computer memory now exceeds a billion bits. Communications technology now easily exceeds a billion bits per second using fiber optics. Storage technology now often exceeds a billion bits in a refrigerator, and soon a breadbox; and so on, and so on. Technology has often exceeded most projections and expectations. The processes of how humans approach a problem and how we explain these problems to machines are closely related. Although current human interfaces and programming languages offer vast improvements, still their development remains in the dark ages. The ability of humans to adequately understand, define, analyze, structure, and utilize a machine model is also quite fundamental in its development. People have a hard time identifying problems and reaching solutions in a normal human environment. When you introduce the handicap of trying to explain a problem using today's programming languages, the task is even more difficult. It is, of course, a major fallacy of computer folklore that machines can, in any real way, replace people. However, this does not mean that the machines or systems of the future will not be as intelligent as people. Fears concerning these developments are both © 17 KNOWLEDGE ENGINEERING justified and unjustified. This seeming paradox portends that we can achieve all of our hopes for machines and for humans as well. There is a dark side to computer advancement that corresponds to the dark side of humans. Despite their benefit to people, intelligent machines often become monsters of the id; they can be used in any possible way for war, invasion of privacy, and human destruction. In this case, machines are only extensions of humans and can be used as their warriors. This issue both leads to and stems from the area of human knowledge ability. Humans must recognize their own frailties, inadequacies, and limitations, and how they affect the technology. This is different from the problem definition issue. Humans might not as yet be adequately equipped to understand all of the forces behind intelligence, thinking, or other human mental processes. We might need to evolve a bit further before we can develop machines that can be as intelligent in as many ways as we are. This is not to say that humans will not design, build, and use these "power tools"; rather, it might take a massive effort of machine power, human-machine interface, and new problem-solving methods just to develop machines that have an average IQ. This book is intended to be an intermediate step between understanding the vast complexities of how KE works and realizing the practical effects it has on how we manage our business affairs. We might not need to know one to understand the other, but it could help. Information Technology in Business Management The technology of the office of the future is painfully emerging. As with children who thought they wanted to be firefighters and now are productivity analysts and systems managers, the office is emerging as a new entity. The people and the technology of the organization are restructuring the organization itself around information. Technology is becoming a survival issue in service industries like banking, brokerage, and sales. If you don't use the available technology to supply the same or better service than your competitor, you cannot win in the 1 marketplace. . . - notes Robert C. Hughes, vice-president of business and office systems marketing for Digital Equipment Corporation. Technology is rearranging organizational schemes from a pyramidal approach to one that appears flat. There has been a decrease in the number, and a change in © 18 TRENDS IN KNOWLEDGE ENGINEERING the type, of management levels in the past few years. In essence, they have been declining, reducing, and shifting power throughout the corporation. Management Styles and Information Technology Matrix management was once highly touted, then almost forgotten. As with most products, timing is the absolute key to success. Matrix management's timing was off by a decade. It will re-emerge as information technology-based systems evolve to provide a communications-based organization. Matrix management was the first step in breaking the bonds of the old production organization. It forced managers and staff to look around and work outside of their closed environment. The matrix organizational chart was fluid, allowing projects, task forces, and internal consultants to operate freely. It was far more open to innovation than a reactive mode of waiting for orders from headquarters. Seeking creative business relationships allows people to excel. This management style is needed more than ever today, with an increasingly independent, educated work force that needs to be challenged. Moreover, as the number of skilled workers declines with the "baby bust," X-type (trustless) managers (Y-managers trust their employees) will not be able to function effectively in the old top-down environment. According to Rosabeth Moss Kanten, matrix management: focuses a new style of management, because in the matrix you cannot order peo- ple around, you cannot command them to do what you want. You don't own the people and own their time, so you have to learn to persuade and once you learn that, you have to learn to sell your ideas. And that, by the way, often means that you need better ideas. Of course, you don't need a formal matrix to do that, just 2 an acknowledgement of cross-functional collaboration. Matrix management probably needs a new name. In this book, I use the term virtual management (VM) to describe a management system that thrives on new technology combined with humanistic concerns. VM, like matrix management, relies more on communication than information. It uses both the technology, such as knowledge networking, and the styles of management that keep people "in touch." VM focuses on the human ergonomic issues, recognizing that all organizations are societies of people. Professional/career and life-style associations In a paper I wrote about five years ago, called "Organizational Marketing (OM)," I projected that future society would fall into two major environments: professional/ © 19 KNOWLEDGE ENGINEERING career associations and life-style associations. If people need a job, a professional/ career association helps them find one. In addition, this association creates new opportunities for faster recognition within the company, especially for aspiring women. It is surprising to talk to men and women about their involvement in these professional associations and to discover how significant they are to these people. The other major association is life-style. In Boulder, Colorado, for example, people are more often identified by their life-style than their occupation. Occupation brings in the money; life-style exercises freedom. Life-style associations run the gamut from sports to religion, from exercise to PC users' groups. Indeed, in today's mobile young communities, such associations often play the role of an extended family. Life-style associations are the "soft side" of humanity. Not just in good deeds, like Boy Scouts, but in terms of art, culture, music, and play—the enjoyment of the world around us and its beauty. These associations have existed for thousands of years, mostly involving games and sports. Many emerge and become popular only to die off, and others like chess, the Olympics, and dice exceed the life span of some religions. The distinction between career and life-style associations is blurred. Certainly, there have always been the gypsies, beatniks, hippies, punkers, and other groups of people who seem to only have a life-style. Others, however, such as Mo Siegel of Celestial Seasonings Teas®, have made careers worth millions capitalizing on life-style, and there are hybrid career/life-style associations, such as the Peace Corps. Business associations are used to further the life-style interests of business associates, and life-style associations often provide opportunities for striking up casual business acquaintances. The ability to create associations of any kind is important, and understanding their significance and their role in the corporate culture of tomorrow is critical for corporate survival. Matrix management, virtual management, and professional associations are corporate cultures at work. They must be assimilated into today's operations and tomorrow's planning for corporate growth. IBM is famous for its corporate culture. It is also famous for its success. There is a correlation. Yet one must always be wary of these "cultural" aspects. Remember that the industrial age had its own style, and with the rapid growth of new technology came new labor unions that were not based on corporations but on specific professional abilities and skills. These associations now number in the thousands and include, for example, the North American Society for Corporate Planning, the International Facility Man- agers Association, and one that I helped start—the International Tele/conferencing Association (IT/CA). These associations flourish because they provide members with the information they need and contribute esteem and awards to people who are often unappreciated within their own company. To many, professional associations are at least as important as the company they work for. Corporations need to © 20 TRENDS IN KNOWLEDGE ENGINEERING understand that the information age will evolve, will possibly become more communications-oriented, and therefore will need a different corporate culture than has previously existed. It will not be enough to use the new technologies to build a better mousetrap. The corporation must use technology to help manage the people who build, sell, and support the mousetrap. Management Sources of Information Computer communications technology helps expand problem-solving capacity by reaching beyond the limitations of time and space. This is a capacity found in the very technological foundations of computers (large memory capacity, high-speed calculation, integrated control functions) in an on-line, real-time system. One of the most advanced problem-solving systems of this kind is the forecasting, evaluating, and warning system. It provides quick discovery of problems in rapidly changing circumstances, forecasts future trends, evaluates the degree of danger created by these problems, and issues a warning when danger appears. To allow use of this technology, some special information management styles will emerge, peddling a variety of sources of information. The principal management sources of information will be in the following areas: • Information providers • Information organizers • Information users • Power brokers Information providers Information providers are those involved in the creation of intellectual values. Information providers will decide what information, or raw data, should be made available to users. They arrange information in a data base for future access. Information organizers Information organizers, or facilitators, will operate computer systems that assist people in at least three ways beyond merely locating information. The systems will be of the following types: • They will be proactive; they can detect a problem before it becomes serious. By analyzing present tendencies and predicting future trends, these systems can present potential alternative solutions. © 21 KNOWLEDGE ENGINEERING • They will discover hitherto unknown problems. Humans tend to find the familiar patterns, and machine systems can consider all possibilities in their calculations. • They will be able to solve problems that are complex and protracted; for example, the contamination problems at Love Canal or a nuclear missile crisis. Information organizers will help build bridges between users unfamiliar with the vast corporate or external warehouses of information. They will manage the acquisition of that information and its condensation into digestible bites for consumption. The growth of information centers as the means of managing, controlling, accessing, reviewing, and moving information will fuel the growth of the organizer's role as a vital interface in corporate management. This manager will have powers similar to a White House adviser. The President must base decisions on the information boiled down and presented by his advisers. The very act of synthesizing information is itself a policy decision. Information users For the information user, new trends are emerging that solve old problems and create very real new ones. The user community is changing rapidly. Computer technology is, on the one hand, fueling this change by providing users with vast amounts of information to increase productivity and, on the other, creating an information "gap" in managing and controlling information because each user is information-autonomous. Users want information, namely their information in their personal computer, and have little regard for corporate integration, protocols, standards, and capacity requirements. The user is a moving target, and in all likelihood is working at home, on the road, or on the move. In response to this phenomenon, some corporate organizations now have personnel who are experts in user needs in both business and life-style. This new breed of user is creating its own data bases, without regard for eventual interconnection to co-workers. The user's principal concern is problem- solving, or decision-making, which involves devising a way to eliminate risks and uncertainties that might stand in the way of accomplishing goals. It also appears that the user does not allow anyone in the organization to stand in the way of acquiring the technological tools necessary to perform the desired task. In many instances, an individual has brought a microcomputer, with some personal software, into the office before the corporation has supplied these tools. Tomorrow's corporation must spot such trends, analyze their impact, and quickly respond to them. User aggressiveness is starting to have a far-reaching impact, not only on the organization but on society in general. The computer is opening new, complex possibilities and creating new social values simultaneously, without regard for © 22 TRENDS IN KNOWLEDGE ENGINEERING economic, political, management, or labor conditions. Thus, the user might be the most complex issue facing management. Power brokers The power brokers, those organizations involved in the technical production of computing, storing, and communicating information, will provide the impetus for the proliferation of telecommunications and information technology. Information generation will be a critical focus in the years to come, whether computer facilities are available for large- or very-large-scale processing. Distributed processing enhances the overall system processing capability by matching it with individual user requirements, whether this takes place at central or remote sites. However, the limitations imposed by different computer operating systems, networks, and dysfunctional software systems will lead to gross shortages in global processing. The winners in this area will be those companies that are able to maintain a high level of processing as well as "broker" unused time to the outside world. Corporate managers relying on global networking have key performance capabilities that transcend corporate boundaries. The ability to network globally will be one of the critical factors in future success. Like Roman legions who were strategically located in distant lands to check and supress uprisings or foreign invasions, networking can electronically bring the equivalent focus at any point. Information gathering can also take place to gain a better understanding of the forces that are needed in any given area. Information gathering points are an important part of this process. Information Processing As previously stated, information providers are the information-age gatherers and farmers of specialized information and data. Information gathering and processing are complex processes. If you gather the wrong information or gather the right information but process it at the wrong time, you end up with the familiar "GIGO": garbage in, garbage out. In market research, most companies gather information in two years, then project by using the growth or decline of the product or market. This is referred to as traditional linear projection-making. There has been a tremendous amount of controversy, however, over strategy versus basic market statistics. It has been argued that market statistics are reflections of "real" sentiments and pressures of the marketplace and/or customer in actual terms. Strategic information, though, has to do with taking the raw data and building the product with it. This question involves © 23 1 0 KNOWLEDGE ENGINEERING more than market demographics because an understanding of market wants is often ahead of the market itself. Strategic planning is the flip side of market statistics. This market consumption of information yields a diminishing value of either strategies or demographics. IBM is both a strategic and a statistical player. It develops in advance the products that can give it a competitive edge. The company is then capable of waiting, due to its size, for market pressure to build to the point where the statistics clearly demonstrate that IBM will probably be able to totally dominate the market. Information organizers also need to be concerned with the processing and packaging of information. Processing information into viable strategies takes on new meaning because of the vast amounts of information being created every day. Over two million words are available every day to a newspaper editor; thus, processing and conveying meaning becomes complex as well. What information is important? And to whom? Do we develop better packaging of information to facilitate its understanding by users? Do we keep it generic or organic, that is, developed without additives? Information Processing Innovation Processing of information in the future might lead to better attempts at innovation. There is some evidence in our own company's development of software, and that of others, of what has been called mindware. This is similar to computer-aided design (CAD) for the mind and ideas. Like product innovation using a CAD system, computer-aided thinking (CAT) might yield the sales strategist new approaches to sales and marketing. These types of systems will, in all likelihood, allow managers to model scenarios and concepts formerly unthinkable for computer simulation. For example, what would be the impact of reduced retirement age requirements on "baby bust" job availability for top-level management positions in recessionary conditions? This example again reflects a balanced approach to information— statistical and strategic. Statistics are needed to give fundamental baseline infor- mation, and strategic players can analyze and give their opinions about impacts or outcomes. Another example is the role IBM will play in the emerging computer-aided communications market. Presently, market forecasters have projected major growth in full-motion video teleconferencing. The subtle issue here is that IBM's product is still-frame video, not full-motion video. IBM understands, among other things, that managers communicate at their desks or in nearby meeting rooms; in other words, managers communicate close to people they know well. There is little need for a picture 30 times a second because most forms of management communication and presentation use transparencies or slides that do not move. IBM developed an innovative product that fit in with their product line and dramatically changed the computer-aided communications market. Although IBM © 24 TRENDS IN KNOWLEDGE ENGINEERING 11 guessed correctly, those companies that put all their eggs in the full-motion video basket did not. This points out how strategic marketing can fail. It has been said that if you think you know what is going on, you are probably wrong. Market forecasts might not truly reflect the market. Remember, even IBM forecast relatively few units of its IBM PC to be sold as executive add-ons to its installed base of larger computer systems. Users, or end users, as they have become known in the computer industry, become more of an enigma every day. Their information needs differ greatly during the course of a day. Their work habits change as frequently as their offices move. Generally speaking, the proliferation of personal computers has occurred as a result of the ability of certain types of software to automate frequent or redundant activities. Reports, including status or action-item updates, and financial and profit-loss statements can now be automated easily by using word processing or spreadsheet software systems. (Automate is used here to describe the mechanical aspects of the function, such as adding or subtracting columns of numbers or correcting spelling and grammar in text.) The expanding base of personal computers has slowed, and will continue to do so until new software systems are developed to assist managers with the less automated, or more complex, aspects of their jobs. Managers and information processing Table 1.1 demonstrates that the majority of a manager's time is spent in commu- nication, followed by telephone conversations, desk work, and thought. It is in the areas of desk work and thought that information technology needs to address issues different from those available today. Simulations and projected outcomes of today's events in the future are typical issues most managers work with. Statistics might not give enough insight to be definitive. The strategy might be too vague for an agreement on the assumptions in the projections. This is the information dilemma faced by most managers: an overload of information that is not filtered well enough to be appropriately used in any definitive way. The role of information technology in creating more data is thus suspect and, in many cases, frustrates managers to the point of blaming the technology for bad decision-making. The other area where information technology is rapidly affecting the end user is in communication. This is discussed elsewhere, but a slightly different issue is at stake here. Information in the form of communication is thought of as electronic text mail. Voice, image, and rapidly approaching video frames are altering the types of information found in many offices today. The packaging of these new types into effective plans and presentations will change the overall nature of management communications in regard to meetings, presentations, and conferences. All four types of information managers discussed will be required to put together such a gathering and presentation of material. The corporations that start gearing up now for these events, from a technological and a management viewpoint, might find © 25 1 2 KNOWLEDGE ENGINEERING TABLE 1.1 Manager's Distribution of Time Current manager's distribution of time • Communicating information transactions Meetings 30% Telephone 20% Travel 20% • Seeking information transactions Desk work 30% Future manager's distribution of time Establishing communications/information transactions Performing people interfaces 40% Meetings Presentations Audio conferencing Video conferencing • Traveling 10% • Seeking information transactions • Performing system interfaces 50% Dictation Telephone-voice mailbox Computer conferencing Viewdata-data systems Decision support systems-assisted "thinking" Computer-assisted retrieval themselves in a powerful position in the information brokerage marketplace of the future. The challenge for KE will be to manipulate these new communication forms into traditional paper-based management functions. New types of software are emerging that have been called visualization systems. In artificial intelligence (AI) terminology, this refers to machines that can relate visually to their environ- ment in a way similar to humans. These systems relate to users as the user desires, rather than as the machine dictates. In designing software systems for end users, a number of issues are present that are generally not considered by most software designers. An entirely new approach to ergonomics (human-machine interface) is needed in designing and using software. A flexible approach to understanding the needs and desires of the changing user is demanded. Software must be able to learn and adapt to the user, rather than the user learn and adapt to vague, uncertain, and complex command key sequences. The trend toward manual-less software and hardware systems with the documen- tation available at any time on-line is important, but unlike most of today's on-line systems, it must be more than putting the paper version into an electronic text file. © 26 TRENDS IN KNOWLEDGE ENGINEERING 1 3 It calls for a reorientation, a rethinking of human communication, and a dynamic interpretation of how a human can relate to a complex machine. (See Foeher and Cross, The Soft Side of Software: A Management Approach to Computer Documentation; John Wiley, 1986.) Multidimensional Management Systems (MMS) The concept of matrix management was introduced by organizational theorists because, among other things, technology was beginning to offer new ways for workers to communicate. As opposed to the traditional management structure, workers could be linked in more than one reporting function. Workers adapted to this management scheme much faster than had been anticipated because it offered the opportunity for excellence in a number of different pursuits rather than in those few pursuits available within the conventional job description and classification. For example, a worker interested in acquiring the skills for managing software programmers has a better chance of doing so in a matrix management system because the structure is more goal-oriented than the traditional control-oriented management systems. Matrix management recognizes and encourages the growing diversity of the new information age work force. VM is the next step in recognizing the role of technology within the corporate sphere. Multidimensional management structure is another hybrid concept in the continuing evolution of management systems. Multidimensional systems are square, round, rectangular, or almost any shape other than pyramidal. Multidimensional systems are not necessarily hierarchical, relational, cross/pointlike, or networklike in their information relationships. There are no roles, just challenges. Multidimensional systems already exist in a number of corporations, and many more are now testing the concept. In multidimensional management systems (MMS), each employee or staff member is recognized as a human data base management system, to use a conventional term. See Figure 1.1 for examples of some emerging management approaches. Each employee might also be called a node, a little less human-sounding, but, nevertheless, an appropriate term for describing the relationship of the person to the network. When management's role was primarily to oversee a worker's actual perfor- mance, the necessity for being close had to do with physical proximity. In information-age environments, this idea is outdated. In our organization, for example, we have had people telecommuting during their entire work life. Occasionally, they have worked in the office—at their option, not the company's. Some have even worked "on the road" like Charles Kuralt. Technology has changed all of the rules. In some cases, technologically-based organizations have no physical proximity to their workers. They have closeness in © 27 KNOWLEDGE ENGINEERING 14 6660 BUS RING OR ANY COMBINATION WAGON WHEEL SPORTS CAR Figure 1.1 An example of multidimensional management systems. another sense, through electronic communications. A similar closeness to the customer can also be established. MMSs are an extension of this concept, yet are appropriate for large organizational systems where staff size numbers in the hundreds or thousands. The MMS organization looks like a network, not a chart. Upper management is more figurative than literal, more dimensional than linear, more spatial than physical. MMSs are innovation-driven rather than management-driven. Because of the relative amount of freedom offered by the communications links, each staff member is allowed the opportunity to explore and interact with anyone at their own time option. Management responsibilities continue in the areas of keeping projects on target © 28 TRENDS IN KNOWLEDGE ENGINEERING 1 5 and maintaining accountability for individual contributions. However, the optimum manager facilitates and does not dictate, only advises and motivates. Already, this concept is strongly emerging, with an increasing number of employees functioning as telecommunications consultants. By definition, their role is advisory to corporate departments. The concept of the information center is almost a user-support function. Many companies are forming other internal user-support groups or clubs to aid users with problems related to the new technologies. In this role, network management takes on a new dimension. It functionally organizes conferences and study groups of people with similar interests but who are in different locations or time zones. In this way, the network is optimized, providing not just information but human resources as well. Managers recognize that the information needed can change daily. The network can act as a multidimensional expert system because most problems have more than one dimension and, therefore, many potential outcomes. Computer-Aided Decision-Making (CAD) The development of computer-aided decision-making follows the basic principles of CAD systems for airplanes, cars, or buildings. It supplements, rather than replaces, human effort in performing a process or task. CAD systems for design and their counterpart, computer-aided manufacturing (CAM) systems, perform a myriad of repetitive tasks at the lightning speed of computers, leaving the most complex and difficult problems to human analysis. It is an understatement to say that CAD systems can be used for solving management problems of nearly all types. This is totally a matter of need. In the discussion of MMS, the role of technology as a management team member takes on new meaning. It will only be a short time before some, if not most, of the functions of management will be automated, supported only by analysts monitoring functions similar to field engineers maintaining a disk drive today. The underlying forces behind this activity are the need to: 1. Optimize decision-making efforts 2. Increase the speed of decision cycles 3. Reduce costs and time 4. Improve decision-making performance 5. Simulate case studies for training 6. Eliminate personnel costs 7. Increase span of control 8. Reduce levels of management 9. Automate personnel skills © 29 16 KNOWLEDGE ENGINEERING The development of computer-aided decision-making systems might be con- sidered complex and infinitely difficult due to the nature of most management positions. However, there are automated telephone sales systems, automatic teller machines, software packages for psychiatric self-diagnosis, and systems for a range of other functions that even a few years ago would have been thought impossible to automate. Many people are wildly enthusiastic about these devices or aids. Children enjoy computer-aided instruction (CAI) because it doesn't scold them, is infinitely patient, and is never tired or bored. You don't have to worry about whether development and implementation of CAD systems will or won't continue; it is only a question of how fast this phenomenon spreads On the bright side, CAD systems eliminate many of the mundane and repetitive functions that keep productivity low and jobs boring. There are three basic areas of CAD systems: inductive, deductive, and intuitive. Models can be easily developed for inductive systems (thermostats, pocket bank tellers, and so on), and systems that perform deductive reasoning are being widely developed today (self-help, brain- storming). Intuitive systems offer the greatest challenge because they attempt to solve problems before they occur, or are proactive in their thinking. Methodologies are needed in order to design these CAD systems. The scope of designing a CAD system consists of the following issues: • Collect relevant information about the problem • Separate data elements into as many distinctly different, yet significant, subcategories • Develop a matrix or network of these components or building blocks • Test components individually and as a group, or simulate actions until systems match a level consistent with human counterparts • Graphically present simulations • Incorporate tests with an overall model • Test in real-life situations to monitor human decisions • Develop procedures for updating and maintaining a system as a regular activity to include outdated, unimportant, or irrelevant variables, as appro- priate The issue of decision-making involvement is a matter of great debate. Most theorists suggest that the expert is the means to the end, that is, you need the doctor to understand the diagnosis. Others conclude a thorough knowledge of the diagnostic results is all that is needed to make recommendations. This is a similar argument to heuristics and computer-modeling processes (i.e., linear programming, regression analysis, queing theory, etc.) versus delphi com- puter conferencing and other humanistic tools. Neither approach is totally correct. As with many decisions, the consequences are long-term rather than immediate. © 30 TRENDS IN KNOWLEDGE ENGINEERING 1 7 Moreover, much of the development of CAD systems has to do with the immedi- ate—today's management activities rather than tomorrow's, where outcomes can be rather vague and dependent on uncontrollable outside variables. CAD systems are seen as a way to do more with less, to become less dependent on human judgments that, for better or worse, are often inconsistent or vague. Much of the research and resulting models are designed primarily for training purposes, where the costs are rapidly rising, instructors are in short supply, and materials are increasingly complex. Also, as the velocity of corporate decision- making increases, static models become quickly outdated. Issues concerning these decisions also become obtuse and difficult to decipher in components for the model to evaluate. Needless to say, CAD systems are arriving with many benefits as well as pitfalls. They will, according to my research, be executive assistants rather than true decision makers. This is due to the fact that most problems require that the decision maker extend beyond the present environment, such as implementing a new system, acquiring a new company, or reorganizing a division. However, as the data base grows, a history of management emerges that can be used as internal "case studies," aiding new managers in their decision-making processing and programmed into the minds of the machine-based executive assistants. Organizational Marketing Organizational marketing (OM) was originally developed in the late 1970s as a marketing strategy for a major Fortune 500 computer manufacturer. At that time, a new approach to marketing goods and services was needed. Even then, computer systems were reaching a level of resistance on the part of users because of the delays by the systems development department or in programming. There were delays of years and cost overruns that are similar to those found in military projects. Users were taking matters into their own hands as well as seeking technological solutions to labor-intensive and time-delayed projects. Certainly there are no easy answers because engineering development has both labor and time dimensions. The notable situation of a low labor component and an infinite time component has produced major innovations throughout history. However, with most corporations, when the time factor goes askew, the only thing they seem to do is to pour in more labor and hope the problem will go away or correct itself. This is seldom the appropriate solution. OM is more of a concept than a methodology. It is a leadership issue. Today, OM is referred to as entrepreneurship, strategic business units, or corporate venture projects. OM has evolved over time into VM, the next generation in management. VM is truly a 1990s corporate management strategy. Just as technology evolves in waves, eras, or generations, so, too, does management. For example, there are people who believe that the computer trend will fade away. Many, if not most, © 31 18 KNOWLEDGE ENGINEERING corporate management systems that were developed in the 1950s still exist today. Like new children, new corporations created today operate under vastly different principles than those formed last year or a decade ago. OM is a system that works with marketing-oriented corporate management. It is based on the simple concept of selling to users, not manipulating them. A few recent books provide examples that are helpful in getting this concept across. In Teleconferencing: Linking People Together, an underlying theme is that this powerful technology is underutilized because the manufacturers of teleconfer- encing hardware and services fail to provide the "driver education" critical to happy users. The analogy is similar to automobile manufacturers selling cars without driver training and thousands of users killing or maiming themselves. You would wonder why people would drive at all People do not teleconference very much today because they don't know how. This will change as users teach themselves how to teleconference. In The Soft Side of Software: A Management Approach to Computer 4 Documentation, the theme is as simple as the title. Software is what the user sees, interacts with, and unfortunately, has to live with and has little or no control over. Sometimes called the "user-defined screen," it allows users to arrange the screen in a way that suits their needs rather than have the software program control the format. Software is beginning to be designed that enhances the intent of the user rather than demands great amounts of mental energy from the user trying to get "help." Software should be strategic in the sense that it develops with, learns about, and provides the user with strategies for its effective use. Case studies, applications, and tactics that propel and compel the user into action rather than stall the user with complicated software systems can each be integrated into the software program. 5 In Intelligent Buildings: Strategies for Technology and Architecture, the approach is to bring together off-the-shelf technologies into a systems approach to office environments. Nearly all of the technology needed for intelligent buildings exists today. Local area networks (LANs), CAD systems (CADs), voice-data digital telephone systems, and energy management are all billion-dollar industries, yet few of them work together in a manageable fashion. It is clearly a situation that is driven by vision of the designer. Developers and facility managers are tired of having technology for which they are responsible go in many directions at once. They expect—no, demand—not only that all of the information and building technologies work together, but also that they be integrated with the chairs and furniture. They have to arrange for all those power, telephone, and computer wires to get from all parts of the building to each individual desk, and are certain that someday the ceiling is going to collapse under the weight of all those cables. These users are taking charge, and they will have their way because they represent billions of dollars of office business. Each situation represents a different user problem and a different technological issue. This suggests that technology is changing toward, rather than away from, user needs. People are slow to adapt, but technology is always changing, often at a rapid pace. It is usually not the user's fault that something is wrong; the user probably © 32 TRENDS IN KNOWLEDGE ENGINEERING 19 does not have enough time to truly understand the usefulness and effectiveness of the technology. OM takes a similar approach to encouraging users to invent new products themselves. OM was conceived as a sales and marketing approach to understanding user needs. OM is user-driven rather than product- or engineering- driven. This concept assumes that users are reluctant to use technology they are unfamiliar with. If corporate technology departments such as office automation, data processing, telecommunications, and so on don't understand how their technology is going to benefit the user, the technology will always fail, except under very compelling circumstances. The sales and marketing of technology is as critical to internal users as it is to the manufacturer. Now you can see why there is a lot of technology sitting around unopened, gathering dust. Technology and design are becoming inseparable. This includes the design of the organization in terms of workers and environments. The advent of the personal computer has enabled many jobs to be performed both automatically like automatic teller machines and autonomously like persons working from home. In terms of the organization, changes in terms of how work is performed and where it is performed will have dramatic effects on how the organization is managed and grows. In this sense, the ability to create environments that are dynamically or virtually adaptable to operate under machine control is becoming essential. Corporations can no longer respond to wildly changing world market conditions on sheer intuition. The extension of machine technology to an organization becomes an organizational issue itself. The machine restructures the company that employs it to seek a new equilibrium from which to act. The machine-human interface sets the stages for modeling simulations from which key management can respond to, rather than be confronted by, disturbances on the economic front. The computer can model and simplify complex conditions into relatively simple "user-friendly" human-machine ergonomics. Therefore, allowing the computer to manage rather than respond is not only viable but will become a strategic weapon in the not too distant future. Virtual Management (VM) New management approaches seem to emerge on a daily basis. At the same time, some experts tell us that keeping an eye on customers and spending time with employees is a time-honored, commonsense, proven management system that works. The problem with both the new and old approaches is that they fail to understand the new employee: the computer The computer has been an employee for over three decades, but few of us recognize this, in spite of the cost justifications when computers took on human tasks. Perhaps we should have given them performance reviews all along. The computer, for some, became an extension of their total personality, © 33