Sapient Agents – Seven Approaches

During the last several years, a large class of computer programs, called “ Intelligent Agents” (IAs), has emerged as a result of intensive research efforts in various countries. There is ongoing discussion how to classify the individual types of intelligent agents and how to distinguish between them and a subclass called “Sapient Agents” (SAs). In this paper, seven approaches are proposed to distinguish SAs from the entire class of IAs. These approaches are proposed mostly in the context of IAs for engineering design, but the results should have more general implications. The proposed approaches are based first on analyzing knowledge representations and the concept of emergence. Considering exploitation versus explorations in design and in evolutionary computations gives further approaches. Next, we consider SAs in the time and domain context, and then in the context of the chaos theory. Finally, initial conclusions and plans for further research are presented.