Models for Integrating Artificial Intelligence Algorithms into Empirical Sociological Research

Abstract

This article analyzes the main discussions surrounding the development of artificial intelligence (AI) algorithms and their penetration into the social sciences. It proposes a typology of models for integrating such algorithms into the professional practices of sociologists. Three models are identified, differing in the types of tasks for which algorithms are used and in the degree of their agency. The article substantiates the thesis that the penetration of AI algorithms into empirical sociology is irreversible and that sociologists’ digital competencies need to be developed. It also provides a brief description of the author’s experience of integrating AI algorithms into the work of a research team. Two levels of digital infrastructure are described: a parser written in Python and an AI agent based on a large language model.