What Is the “Human” in Human-Centered AI? Restoring the Aesthetic Subject

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Human-centered artificial intelligence (HCAI) seeks to orient AI toward human values, autonomy, oversight, safety, and meaningful control. Yet the term “human-centered” leaves a prior question unresolved: what conception of the human is being centered? Much HCAI discourse approaches the human as user, decision-maker, rights-holder, stakeholder, or bearer of preferences and values. These conceptions are indispensable but incomplete. This article argues that HCAI must also recognize the human as an aesthetic subject: a being whose perception, attention, imagination, interpretation, and judgment are formed through experience. As recommendation, personalization, generative AI, and predictive systems increasingly organize the conditions under which people encounter and evaluate the world, AI can participate in forming rather than merely serving human preferences. The article develops the concept of aesthetic agency to describe the human capacity to participate meaningfully in the formation of one’s perceptual, imaginative, interpretive, and evaluative relation to the world. It distinguishes choice from judgment, preference from taste, generation from imagination, and optimization from experience. On this basis, it proposes an aesthetically expanded HCAI organized around openness, plurality, reflectivity, and transformability. To center the human is not only to preserve control over AI, but to preserve the capacity to perceive, imagine, judge, and become otherwise.