Sociocultural recursion in the context of actor-network interaction with generative artificial intelligence

Technologies of generative artificial intelligence (hereinafter referred to as GenAI) are becoming an integral part of society, taking part in various socio-cultural and socio-economic spheres, and therefore they actualise the problem of socio-cultural reproduction. Despite the active development of this research vector in terms of generative synthesis effects and risks, there is no detailed conceptualisation of transition processes between the current and predicted states. The purpose of the article is to reveal the essence of the process of forming socio-cultural distortions when interacting with the GenAI. The methodological basis is the actor-network theory. The study reveals the actor structure of interaction with the GenAI as part of the composition of social and generative actors and data array mediating their connection. It is argued that the resulting nature of socio-generative interaction is determined by the social predestination of the data array, which we propose to distinguish in four variants. Moreover, these variants set four types of relationship between the actors, and in the end, socio-cultural meaning-making occurs. The process of transition including the direct one (in which the information-objective result of social-generative interaction transfers to its socially subjective representation) and the revers one (in which the generative-subjective result transfers to its socially objective representation) are conceptualised. It is concluded that this process is a recursive cycle of distorting reproduction of the socio-cultural system. The results contribute to the conceptualisation of the phenomenon of the AI and its role in social systems, complement the discussion regarding the likely effects and risks for society and can serve as a basis for developing regulatory solutions in various areas of the GenAI use.