CONCEPT
Post-Epistemic Environment
The information condition that generative AI threatens to produce: a public sphere saturated with synthetic content indistinguishable from authentic content, in which the signals that previously allowed citizens to evaluate information—effort, expertise, institutional backing, authentic human experience—become systematically unreliable.
Democracy requires a shared information environment. This is not a preference but a structural prerequisite: democratic deliberation depends on citizens sharing enough common ground—not agreement on conclusions, but agreement on basic evidentiary standards—to engage in productive disagreement.
Al Gore traced the erosion of this common ground from television through social media in his
Assault on Reason (2007), but his framework anticipated the deeper crisis that generative AI threatens to produce. Previous disinformation required human effort—someone had to write the misleading article, create the doctored image, compose the inflammatory post. The effort imposed a natural limit on volume and left traces that made detection possible. Generative AI removes both constraints simultaneously: unlimited synthetic content, indistinguishable from authentic content, at near-zero marginal cost. When AI can produce “
confident wrongness dressed in good prose”—plausible, articulate, well-structured content that is factually incorrect—the burden of verification shifts entirely to the consumer, in precisely the information environment where citizens already lack