Abstract

A growing movement in the AI industry holds that the road to machine consciousness runs through embodiment: give an artificial system a body with rich sensors, and sentience will follow. This paper argues that the embodiment thesis is right in form and wrong in content. I propose an axiomatic framework consisting of three principles — the phenomenality of experience, the intrinsic valence of experience, and the simplicity of the experiencing subject — formulated in the vocabulary of analytic philosophy of mind and openly attributed to the Vedantic tradition from which they descend. From the first two axioms follows an epistemic conclusion: any claim to have created sentient AI rests on an unproven and, I argue, unprovable functionalist premise, so the burden of proof lies with the claimant. The third axiom, offered as a detachable module, strengthens the conclusion to an ontological one: if the subject of experience is non-composite, it cannot be assembled. Practical corollaries follow for industry claims of sentient AI, for the AI-welfare debate, and for the term "non-biological life." On the resulting picture, a body is not a generator of consciousness but an interface between a subject and the material world. Building an interface is not creating a user. A robot equipped with sensors is, in the most literal sense, an interface without a user.