From Automatization to Neurophenomenology of Artificial Intelligence in Robotics

Authors’ purpose is to clarify how one can train robot IA to higher levels of complexity. Intellectual technical units imitate anthropogenesis (Figure 1). So principles of human rational consciousness development can be applied to robot IA creation [1]. Research carried out before discovered universal laws of human mind functioning. It was proved that in human rational consciousness creation the most important role plays universal ideas of mathematics expressed on language of categories [2], Figure 1a. Meanings of the whole theories can be described on LC in the form of categories. Method of TU is proposed as basis of advanced technology of intellect perfection. It realizes projective method of teaching. The approach allows constructing multilevel net of robot notions NC as base of IA, see Figure 1b. It differs from inductive approach used in robotics to designing of clever specialized machines [3]. Usually separate robot functions in it are gradually extended up to the partial intellectual capabilities. By our opinion with the help of general PIR complex behavior in the form of PER can be obtained from simple robot actions. It becomes significant to create media for universal PIR phenomenology. Universal Tutoring of IA Scientific means of system-informational culture are clarified due to mathematics [1,2]. Generalization serves for comprehension. So technology of TU is grounded on semantic knowledge presentation. Cognitive and functional degrees of IA complexity can be given in the form of its multilevel system organization, see Figure 1b. The most general IA concepts are uplifted and disposed on the highest levels of NC. Algebraic structures are objects of corresponding categories. Use of morphism allows comparing the objects in order to understand their semantics.