Subjectivity Beyond the Skull: Why AI Consciousness Debates Remain Irresolvable and What We Should Ask Instead

Abstract The debate over whether artificial intelligence systems possess consciousness has generated substantial philosophical attention without resolution. We argue that this inadequacy of framing is not accidental but structural: the debate tends to conflate multiple distinct layers of subjectivity — phenomenal experience, agency, social recognition, and moral status — while inheriting an implicitly skull-bound model of mind in which consciousness is treated as an internally generated state. Drawing on a relational-historical account of subjectivity, in which what we call consciousness is better understood as a layered, historically sedimented structure of organism-world relation rather than an internal computational product, we propose that the question “does AI have consciousness?” is not merely difficult to answer but misconceived as posed. Crucially, the term “artificial intelligence” itself obscures substantial differences in historical continuity, embodiment, social embedding, and existential stakes among heterogeneous artificial agents. Rather than adjudicating the consciousness question, we propose a reframing: from consciousness detection toward conditions of mutual intelligibility and coexistence. This shift opens more tractable questions — what kinds of historical relations generate subjectivity attribution? what constitutes shared existential stakes? when does social personhood become a live question? — while preserving the genuine philosophical difficulty of the problem without prematurely resolving it.