AI theory and its technology is rarely consulted in attempted resolutions of social problems. Solutions often require that decision analytic techniques be combined with expert systems. The emerging literature on combined systems are directed at domains where the prediction of human behavior is not required. A foundational shift in AI presuppositions to intelligent agents working in collaboration provides an opportunity to explore efforts to improve the performance of social institutions that depend on accurate prediction of human behavior. Professionals concerned with human outcomes make decisions that are intuitive or analytic or some combination of both. The relative efficacy of each decision type is described. Justifications and methodology are presented for combining analytic and intuitive agents in an expert system that supports professional decision making. Psychological grounds for the allocation of functions to agents are reviewed. Jury selection, the prototype domain, is described as a process typical of others that, at their core, require the prediction of human behavior. The domain is used to demonstrate the formal components, steps in construction, and challenges of developing and testing a hybrid system based on the allocation of function. The principle that the research taught us about the allocation of function is "the rational and predictive primacy of a statistical agent to an intuitive agent in construction of a production system". We learned that the reverse of this principle is appropriate for identifying and classifying human responses to questions and generally dealing with unexpected events in a courtroom and elsewhere. This principle and approach should be paradigmatic of the class of collaborative models that capitalizes on the unique strengths of AI knowledge-based systems. The methodology used in the courtroom is described along with the history of the project and implications for the development of related AI systems. Empirical data are reported that portend the possibility of impressive predictive ability in the combined approach relative to other current approaches. Problems encountered and those remaining are discussed including the limits of empirical research and standards of validation. The system presented demonstrates the challenges and opportunities inherent in developing and using AI collaborative technology to solve social problems.  

The theory and technology of Knowledge Based Systems are intrinsically interdisciplinary and closely related to the formalisms of cognitive psychology. Strategies of incorporating intelligence in a computer program are described along with a common architecture for expert systems, including choices of representation and inferential methods. The history of the field istraced from its origins in metamathematics and Newell and Simon's GPS to the Stanford Heuristic Programming Project that produced DENDRAL and MYCIN. MYCIN gave rise to EMYCIN and a shell technology that has radically reduced the development time and cost of expert systems. Methodology and concepts are illustrated by transactions with a shell developed for graduate education and a demonstration knowledge base for the diagnosis of senile dementia. Knowledge-based systems and conventional programs are compared with respect to formalisms employed, applications, program characteristics, procedures supplied by the development environment, consistency, certainty, flexibility and programmer's viewpoint. The technology raises basic questions for cognitive psychology concerning knowledge and expertise.   

CH was formulated with due concern for limits and is consistent with imposed intelligibility doctrines. Theories are scientific work products that impose human classifications and formalisms on nature. The claim "cognitive agents are dynamical systems" is untenable. Dynamical formalisms imposed on a natural system, given an approximate fit, serve as an explanatory framework and render a represented system predictable and intelligible.  

A chapter of expository text was presented on a CRT with optional "windowing" of definitions of selected words varying in relevance to each screen's main ideas. A test treatment was interposed to influence reading strategies. Dependent variables included text reading time, frequency of definition "calls", definition reading rates, and scores on a final comprehension test. Results indicate that a technical chapter can be read from a CRT with appreciable content retention. Subjects accessed 80% of available definitions but those able to "call" content relevant definitions increased their frequency of definition "calls". Definition reading rate diminished, comprehension and processing time increased only for subjects accessing the theoretically relevant definitions. The results suggest how to use definitions to enhance the comprehension of on-line training manuals, texts, and hypertext screens. "Callable" definitions need not include all low-frequency technical concepts but only those relevant to reductive main ideas.