I propose a two-dimensional taxonomy to classify human-machine perceptual systems along (1). Embodiment, from non-embodied static perception to embodied sensorimotor coupling, and (2) Qualitative Experience, from purely functional mapping to systems that could support subjective, affective states. The resulting four quadrants (A–D) capture existing architectures such as conventional CNNs, active vision robots, and theoretical qualia-inclusive models. I discuss how the taxonomy informs model inte…
Read moreI propose a two-dimensional taxonomy to classify human-machine perceptual systems along (1). Embodiment, from non-embodied static perception to embodied sensorimotor coupling, and (2) Qualitative Experience, from purely functional mapping to systems that could support subjective, affective states. The resulting four quadrants (A–D) capture existing architectures such as conventional CNNs, active vision robots, and theoretical qualia-inclusive models. I discuss how the taxonomy informs model interpretability, highlights opacity trade-offs, and suggests evaluation protocols that jointly assess functional adequacy and experiential fidelity. Future work will explore the feasibility of Quadrants B and D by investigating frameworks for integrating affective modules and embodied interaction into current cognitive architectures.
