Emergent Novelty in Human-AI Dyadic Systems: A Theoretical Framework and Preliminary Evidence

Abstract

This paper introduces Emergent Novelty as a proposed measurable property of sustained human–AI dyadic systems: output that may arise from the interaction dynamics of a particular human–AI pairing rather than either participant considered alone.

It proposes a three-question Novelty Attribution Test, introduces the Emergent Novelty Rate as a calculable metric, and reports preliminary findings from one longitudinal case study. The evidence remains limited. The framework, its attribution method, and the reported rate require independent replication, contrary cases, and further empirical validation.