Consciousness and AI

Abstract

The primary aim of this article is to establish the presence and/or absence of consciousness in artificial intelli-gence (AI). To achieve this, the authors address the following research objectives: to characterize existing ap-proaches to understanding the nature of AI, to identify their strengths and weaknesses, to determine which ap-proach allows for establishing the presence and/or absence of consciousness in AI, and to identify social and humanitarian factors that hinder an adequate assessment of AI. The object of this study is AI, and the subject is the problem of whether it possesses consciousness. To solve this problem, the authors draw on the behaviorist approach (the Turing test), which has been in use until recently, and the relatively new functionalist approach. Using the method of critical analysis, the authors substantiate the inadequacy of applying the behaviorist ap-proach in assessing the presence and/or absence of consciousness in AI. This approach is based on the out-dated notion of consciousness as an exclusive property of human nature and does not allow for an objective assessment of the presence of consciousness in AI. In contrast to the behaviorist approach, the functionalist approach allows for subjective experience without a permanent “self” and provides grounds for believing that not only humans but also artificial systems can possess consciousness. In conclusion, it is argued that the ap-plication of the functionalist approach will help overcome society’s cautious attitude toward AI and harness its potential for the progressive development of humanity.