Gamboa, Dirk: Beyond Evaluation: World Notions, Normativity, and the Structural Simulation of Ethics in AI Systems

_Zenodo Repository_. 2026This paper develops the Ethics-by-World-Notions framework as an approach toward explicit evaluative cognition within artificial systems. Contemporary AI architectures primarily operate through semantic prediction and statistical approximation without explicitly representing the evaluative structures underlying normative behavior. The paper argues that transparent normative reasoning instead requires explicit world notions through which propositions, actions, and goals be