Abstract
In the first part, the author discusses the methodological status of interdisciplinary research. Some general issues may be problematized within various disciplines; each of them only examines a certain aspect of the issue. Philosophical categories represent some ideas par excellence, while the terms used by the specific sciences represent models of some of their aspects. The second part recalls the most important problems signaled by engineers and programmers of machines, which are to operate in an external, changing environment, realizing autonomously the objectives set for them. In parts three and four, the author attempts to specify the essential operations that the decision making machine performs. The nature of these operations should be adapted to the requirement that the machine should focus on what is important for the task it performs.