Artificial intelligence, organizational coordination, and productivity heterogeneity: An evolutionary model of structural change

This article examines whether artificial intelligence (AI) may constitute a new sociotechnical paradigm by altering the ontology of productive knowledge and, consequently, the microfoundations of firm theory. Building on the evolutionary and neo-Schumpeterian tradition, it argues that sociotechnical paradigms can be understood as historically specific ontologies of knowledge, shaping how knowledge is generated, validated, and incorporated into economic activity. Within this framework, the article develops a conceptual and analytical extension of the microfoundations of the firm by incorporating algorithmic mediation into three core dimensions: learning, routines, and dynamic capabilities. The analysis suggests that AI may introduce a conditional and heterogeneous shift in the nature of productive knowledge, characterized by greater codifiability, scalability, and partial autonomy of inference processes. These transformations may modify learning mechanisms through data-driven model training, may reconfigure routines via hybrid human–algorithmic decision structures, and may extend dynamic capabilities toward the orchestration of distributed cognitive resources. The results suggest that AI does not replace existing microfoundations, but alters the conditions under which they operate, leading to distinct organizational regimes depending on task characteristics, data quality, and governance constraints. The article contributes by linking changes in knowledge ontology to transformations in firm organization.