Automating Intelligent Systems: Technological Determinism and Substantivism

Abstract

Artificial Intelligence has become so firmly embedded in our lives that its direct influence on shaping the world of the future is inevitable. However, it has taken time for a constructive approach to risk prevention and regulation of technologies at all stages of their life cycle to gradually emerge alongside theoretical speculation about «machine uprising» and other threats to humanity. The subject of special attention is the so-called automated artificial systems, the regulation of which is still limited by normative and technical requirements. The peculiarity of this approach is the conviction of its proponents in the truth of technological determinism, for which «technology» is value neutral. The prevention of ethical risks from the perspective of this approach is practically impossible because regulatory issues are only concerned with the functional characteristics and operational violations of a particular system. This article contrasts technological determinism with technological substantivism, for which «technology» has an independent ethical value, regardless of its instrumental use. The ethical evaluation based on it consists in the procedure of regular correlation of social «good» and «reliability» of the system. The development of a methodology for such a correlation procedure requires special competences that distinguish a new professional field — ethics in the field of AI.