On the Task Approachin Artificial Intelligenceand Cognitive Sciences

This work continues a series of publications on the task approach in artificial intelligence. As noted earlier, the agent-based approach described in the monograph by Stuart Russell and Pieter Norwig “Artificial Intelligence. The Modern Approach", may be more argumentatively presented within the framework of the task approach. This paper will show that not only the problems of the bases of mathematics and artificial intelligence, but also many cognitive functions performed by humans and analyzed in cognitive sciences, can also be described and studied within the framework of the task approach. In particular, this paper shows that the analogue of the concept of task in cognitive sciences is the concept of goal and that the Functional Systems Theory (FST), which describes purposeful behavior, can be presented as the brain's solution of tasks to achieve goals and satisfaction of needs. It gives the chance to compare directly the tasks of artificial intellect with natural cognitive processes and, thereby, to reveal the list of those tasks of "natural" intellect and schemes of their solution which can be successfully used for the solution of artificial intelligence tasks.

artificial intelligence, general artificial intelligence (AGI), cognitive science, agent-based approach, task approach, purposefulness activity, purpose and task, Functioning System Theory (FST), decision making, rational agent, rational agent modeling