Mass media depends on the creation and circulation of types of people. Its effort to attract a particular consumer through messages addressed to a mass audience produces an aesthetic gesture caught between the particular and the “general,” one grounded in assumptions about social and cultural types. Generative AI intensifies this dynamic, circulating generic content while optimizing it for personalization. In this talk I take a semiotic approach to the type, showing how its construction is both citational and performative. Rather than simply representing established categories, performances of types in visual culture – from Francis Galton’s composite portraits to stock photography and generative AI – reveal the type’s fictionality and its capacity to undermine itself.

Drawing on Judith Butler’s account of performativity, I show how the generic type approximates the real while also exposing the conditions of its own performance. An AI-generated image of a “college student” tries to pass as something recognizable, yet in doing so lays bare the social and cultural cues that make the performance legible. The type can thus sediment clichés and perpetuate violence, but it can also expose the arbitrariness of its construction.