The convergence paradox: a constructivist account of perceived template convergence and originative agency under generative artificial intelligence

Abstract

Generative artificial intelligence (AI) has raised a widely shared concern that human cognitive and creative output is drifting toward common templates. The terms used to describe this drift, including algorithmic monoculture, outcome homogenization, and the loss of collective diversity, all characterize systems, populations, or model outputs. Psychology has not supplied a matching construct at the level of the individual, that is, a way to describe how a person perceives such convergence and how that perception reshapes agency. This article introduces and analyzes perceived template convergence (PTC), defined as a person’s appraisal that the construals available to her through generative AI are collapsing toward a shared template. Working from personal construct theory, the agentic strand of social cognitive theory, and research on meaningful work, the analysis argues that PTC is psychologically two-sided. The appraisal that raises a person’s confidence in producing competent output through AI can also weaken her sense of authorship and meaning, steering effort toward imitative production and away from the more demanding work of origination. We call this structure the convergence paradox, set it out as a propositional model with two mediators and two boundary conditions, distinguish PTC from neighboring constructs, and propose a measurement agenda. The contribution recasts algorithmic homogenization as a question about human construing rather than about outputs.