Current Personal Knowledge Management (PKM) systems are designed around a single primitive: the note — a unit of captured knowledge. While effective at storage and retrieval, these systems fail to address a distinct and underexplored problem: the accumulation of undetected epistemic contradictions over time. In high-information environments, individuals regularly absorb conflicting beliefs without awareness, and frequently regress to previously defeated positions due to the absence of longitudinal belief tracking. We propose Antinomia, a PKM architecture whose fundamental unit is not the note but the tension — a formalized contradiction between two held beliefs. Antinomia introduces five memory layers, an AI-driven Contradiction Hunter, a belief lifecycle model that tracks not only current knowledge but its full epistemic history, and an Epistemic Network layer enabling structured knowledge sharing between vaults. We argue this architecture produces a qualitatively different cognitive artifact: not an archive of what the user knows, but a dynamic map of how their thinking evolves under pressure — both individually and collectively.
Giuseppe De Martino (Sun,) studied this question.