The Importance of Kant’s Philosophy of Mind for Contemporary Research in Artificial Intelligence

Since its inception, the artificial intelligence program has relied on a positivistic, anti-psychological philosophical paradigm, in which a purely physicalistic description of thinking processes assumed their adequate modeling using logical machines relevant to the tasks and goals, such as Turing (1960s-70s). Optimistic expectations of positive results immediately ran into both technical difficulties and purely conceptual difficulties. However, when the urgent problem of philosophical revision of the basic AI paradigm arose, Kant’s theory of consciousness and thinking was not seriously considered and was criticized in the 1990s. Since the 2000s, we have seen impressive successes in the use of artificial neural networks with deep learning architecture in the field of modeling thinking and complex biological processes. It seemed that the main goal of the AI program - achieving strong AI - was just a matter of time. But the direct implementation of the connectionism concept in working with large volumes of associative and fuzzy arrays of information turned out to be generally ineffective in the field of representing the intellectual abilities of consciousness, especially in the representation of high-level knowledge and precise processing of symbolic information, i.e. higher cognitive abilities. At the same time, some AI specialists and cognitive philosophers turned to Kant’s philosophy of consciousness, which embodied such a transcendental organization of the macroarchitecture of an intellectual system that has an active cognitive activity, but does not correspond to modern ideas about the various mechanisms for processing input and output data in a cognitive system. Such cognition is fundamentally active, since it is a product of the synthesis of the ability of productive imagination. To identify this macroarchitecture, the Kantian transcendental method is used, which consists in the fact that the transcendental architecture of any consciousness is not created as a result of empirical studies of human intellectual abilities, the functioning of brain processes or the achievements of evolutionary biology, but is constructed based on the a priori conditions of the very possibility of its existence. This Kantian method aims to reveal an a priori structure of consciousness that is isomorphic to any rationally knowing subject. The study examines what Kant’s philosophy has to offer AI and cognitive science.