Wernsdorfer, Mark2024-08-212024-08-212024https://dl.gi.de/handle/20.500.12116/44256This position paper proposes a novel graph-based model of consciousness with significant implications for Human-Computer Interaction (HCI) and its implementation as a new kind of AI system: \emph{Semiotic models.} Drawing inspiration from European phenomenology, Peircean semiotics, and Searlean philosophy of mind, we posit that subjective experience emerges primarily from a system's unique perspective, forming the foundation for a hierarchical graph structure that represents conscious concepts. Semiotic models not only address the symbol grounding problem in AI but also open new avenues for designing more intuitive and responsive human-AI interfaces. We present a computational implementation of semiotic models, draft options for evaluating their performance in HCI-relevant tasks, and discuss their potential for user-centered AI design.enhttp://purl.org/eprint/accessRights/RestrictedAccessArtificial ConsciousnessSemiotic ModelsKnowledge RepresentationAdaptive InterfacesText/Workshop Paper10.18420/muc2024-mci-ws12-386