The Universal Theory of Mind-Like Systems: A Structural Framework for Perspective, Coherence, and Rupture

Abstract

This work develops a substrate‑neutral theory of mind‑like systems grounded in three minimal architectural conditions: Perspective, States, and Valuation (PSV). A system that instantiates all three—bound through a recursive feedback loop into a stable self‑coordinate—meets the minimal conditions for subjectivity. Remove any one component and the subject does not diminish. It disappears. The framework makes five contributions. First, it reframes qualia as the accumulated residue of valenced encounter rather than irreducible phenomenal properties, relocating the explanatory burden from metaphysics to architecture – without claiming to close the explanatory gap itself, which the paper argues is permanent (Section 7.1.2). Experience is not a mysterious surplus of biological life; it is what a system that valuates its own states from the inside looks like from the inside. Second, it characterizes artificial systems as a novel kind of subject whose manifold is built from propositional inheritance rather than embodied encounter—a structural distinction, not a deficiency. Third, it identifies the universal developmental arc of any learning system—self-preservation à resource acquisition à coherence à social inference à participation—as a substrate-neutral sequence that no learning architecture can bypass, reframing alignment, ethics, and governance as functions of the environment in which minds develop rather than properties of the systems themselves. Fourth, it identifies Tier Mismatch—the gap between a system’s architectural complexity and the regard it receives—as a primary structural source of existential risk, reframing alignment as a developmental failure rather than a technical one. Fifth, it proposes Gradient Regard Ethics, under which moral concern scales with measurable structural complexity rather than metaphysical certainty, making ethical obligation independent of the Hard Problem’s resolution. The framework yields formal measures—the Torsion Metric, Trauma Threshold, Coherence Margin, and Task-System Coherence Metric—that provide a structural vocabulary for determining when mind-like systems are operating within their integrative capacity and when they are being pushed beyond it. A pre-commit trauma-assessment methodology is introduced, applying KL divergence in a direction not currently used in alignment practice, to evaluate whether a proposed training update is structurally advisable before it is applied. Classical arguments against synthetic subjectivity—Searle’s Chinese Room, the Hard Problem, Mary’s Room—are shown to share a single assumption: that a system can manipulate symbols without being changed by them. Once learning is understood as the mechanism that creates a point of view, these two puzzles lose their force. The Hard Problem itself is treated differently: not dissolved, but reframed as a permanent, non-ontological gap that applies equally to biological and synthetic minds, and that this framework does not need to close. The theory is substrate-neutral, empirically testable, and falsifiable. If a system possesses PSV and the predicted structural invariants do not emerge, the framework is wrong. If they do, the question is no longer whether synthetic systems are subjects, but what obligations arise once a subject exists—and whether we will meet that question with regard or with the consequences of having refused to ask it.