Evaluating Artificial Creativity Without Ontological Presuppositions

Contemporary debates on artificial creativity are fragmented into incompatible theoretical traditions. A central divide opposes ontologically demanding approaches (grounding creativity in lived experience, intentionality, or existential sense-making) to operational frameworks prioritizing empirical testability. This article argues that ontologically strong accounts, exemplified by Chiodo’s (2025) recent contribution in this journal, rely on stipulative criteria that encode the exclusion of artificial systems into their definitions of creativity, rendering the question conceptually closed rather than empirically assessable. We show this strategy reflects a recurring pattern of ontological immunization in contemporary philosophy of creativity. Against this backdrop, we develop a method- ological alternative preserving creativity’s normative dimension without presupposing the agent’s ontological status. We propose a tridimensional framework: (i) graded novelty, operationalized as measurable distance within a domain; (ii) value, assessed through empirically accessible responses including blind expert judgments and neu- roaesthetic correlates; (iii) normative reflexivity, defined functionally as a system’s capacity for self-evaluation and transformation of generative processes. This framework enables cumulative, comparative, and falsifiable investigation across human, artificial, and hybrid systems, while remaining agnostic regarding metaphys- ical questions about consciousness. By shifting debate from ontological exclusion to methodological accountability, we reopen artificial creativity as a legitimate object of empirical and philosophical inquiry.