A human element in a human-machine hybrid of artificial intelligence

The article is devoted to the question of what happens to the status of a human being in the modern period of artificial intelligence technology (hereinafter — AI) dominance. Anthropocentrism is treated as an insufficient conceptual means for the identification of the real position of a human, because it leads to the fixation of simple instrumentality behind AI systems. Such a distribution of roles between humans and technology does not correspond to the real complexity of the functioning and interaction of humans and AI at the present stage. AI, by virtue of its construction, has a certain degree of autonomy, opacity and unpredictability, which allows us to generally talk about it as an agent. However, if we allow the expansion of the field of morality as a result of its reorientation from the subject of action to the object and its moral state’s significant changes, then artificial agents can be regarded as moral agents (based on Luciano Floridi’s Information ethics project).

However, these agents are designed and trained by humans based on data that is produced, collected, and marked up by humans again. The reality of technology production in its social aspect is so complex that, as a result, it is not possible to take into account the separate contribution of all agents (human and non-human) to its functioning. Therefore, the concept of a sociotechnical multi-agent system is used to designate this situation. It is characterized by different levels of agency: causal, moral and moral with the possibility of responsibility. The human in the sociotechnical system is a source of moral agency and responsibility due to the fact that they are its essential properties. The further direction of the research is connected with the question of how the problem of eliminating the responsibility gap in the realm of the sociotechnical system should be solved.