Sensualism and a prioriism as the basis for the evolution of artificial intelligence
р.
р.
Makhamatov, Tahir Makhamatovich Doctor of Philosophy, Professor, Department of Humanities of the Financial University under the Government of the Russian Federation, Professor, Diplomatic Academy of the Ministry of Foreign Affairs of the Russian Federation, Moscow, Russia makhamatov.tair@mail.ru
Timur, Tahirovich Mahamatov Ph.D., Associate Professor, Department of Sociology, History, & Philosophy, Financial University under the Government of the Russian Federation, Moscow, Russia timour,makhamatov@gmail.com
Abstract The presentation substantiates the thesis that the development of artificial intelligence technology is closely linked not only with discoveries in the field of natural science, anthropology and medicine, but also with achievements in the field of philosophy of cognition and cognitive sciences. Authors offer the thesis with a philosophical and epistemological analysis of the problems of improving the neural network as the core of modern artificial intelligence and concludes that the principles of functioning of the neural network correspond to the principles of sensualism of J. Locke, I. Kant's a priori. The results of the comparative study allowed the authors to come to the following conclusion: the improvement of the abilities of the neural network studied by S. Haikin, S. Russell, P. Norvig ("evidence of the answer", "classification of images" and "reliability of the decision") is possible based on the epistemological ideas of J. Locke, the use of Kantian principles ("synthetic unity of apperception", "I think") and the search for an algorithm for the formation by a neural network of the ability to create antinomies in artificial intelligence.