Beyond Homogenization: Cultural Entropy as a Structural Property of Algorithmic Media

Algorithmic media systems are widely observed to produce a paradoxical combination of expanding cultural production and contracting symbolic diversity, but existing concepts, including filter bubbles, algorithmic homogenization, and platform power, describe this phenomenon at the level of outputs or institutional relations rather than theorizing its generative mechanism. This paper develops the concept of cultural entropy: a structural reduction in the space of symbolic differences that can achieve visibility, persistence, and distinctiveness within algorithmic media environments. The term is used heuristically, identifying parallels with thermodynamic and informational entropy in their directionality, dissipation, and limited reversibility, rather than claiming that culture is governed by the same formal laws. Distinguishing cultural entropy from homogenization, a contingent and output-level convergence, the paper argues that the entropic dynamics of algorithmic media are non-contingent, emerging from the fusion of algorithmic and economic rationality within a single operational infrastructure, a configuration without precedent in twentieth-century mass media's separation of economic and editorial logics. Drawing on McLuhan, Manovich, Baudrillard, Postman, and Luhmann, the paper develops a theoretical account of the mechanism, comprising performative selection, self-reinforcing feedback, and the formalization of external difference, and examines its operation through the case of Meta. The paper concludes by examining generative artificial intelligence as a recursive intensifier of these dynamics, in which systems trained on already-selected cultural production generate further content according to the same entropic logic, narrowing the distance between cultural production and algorithmic selection.