These research reports and discussions are concerned with the information processing activity that underlies intelligent behavior in human beings and computers. In the introduction to part one on artificial intelligence, we present our understanding and interpretation of the goal of this research. We have selected reports of research efforts which we feel outdistance all others in advancement toward this goal. Such a criterion, as we see it, gives high priority to a particular brand research, loosely labeled "cognitive models." An opposing school of thought, sometimes called "neural cybernetics" or "self organizing systems," has intrinsic fascination and has produced a considerable number of particular projects. Neural cybernetics approaches the problem of designing intelligent machines by postulating a large number of very simple information processing elements, arranged in a random or organize network, and certain processes for facilitating or inhibiting their activity. Cognitive model builders take a much more macroscopic approach, using highly complex information processing mechanisms as the basis of their designs. They believe that intelligence performs performance by a machine is in and difficult enough to achieve without "starting from scratch," and so they build it into their systems as much complexity of information processing as they are able to understand and communicate to a computer (using their programming techniques). the cognitive models approach has led to tangible progress (displacement toward the ultimate goal) in the field of Art official intelligence, while the progress to date dinner out in the neural cybernetics approach is barely discernible. On this basis, we feel that there is reason for our bias in favor of cognitive models, though of course there are other dimensions along which to evaluate the research.