Algorithmic Experiential Structural Platonism: The Philosophical Stance of Kolmogorov Theory (KT)

Reality is one experiential ground whose structural face is mathematics — and agents are the persistent programs that carve out time, valence, and physics from within it.

Kolmogorov Theory (KT) starts from the one thing you can't argue away: you are having an experience right now. Rather than trying to derive experience from matter or computation, KT takes experience as the bedrock and asks what structure it must have. The answer is mathematics — not as a separate Platonic realm, but as the structural aspect of that same experiential base. Think of it as two faces of one coin: the intrinsic face is pure, undifferentiated awareness; the structural face is the relational, mathematical organization of that awareness. This is what the paper calls "two-aspect monism," and it's the core move that distinguishes KT from both standard physicalism and from Tegmark's Mathematical Universe Hypothesis (which posits math alone, with no experiential ground).

Agents — biological or artificial — are persistent, self-regulating programs embedded in this structured base. They carry three functional modules: a Modeling Engine (M) that compresses sensory history into a world-model, an Objective Function (O) that scores futures, and a Planning Engine (P) that selects actions. Valence — pleasure and pain — is the output of O applied to the agent's model, not a property of the base itself. This is a sharp claim: the base is non-valenced, impersonal, and subject-free. Suffering and joy are structural properties of agents, not of reality's ground. This connects directly to reinforcement learning and dopaminergic reward-prediction-error data, giving the framework empirical traction.

Time, in this picture, is not fundamental. The paper develops a "tile-universe" — a static, global mathematical configuration governed by local matching rules, rich enough to encode Turing-universal computation. What we experience as the flow of time is an agent-relative compressive coordinate: the slicing of that static structure that best predicts the next state. This is close to Barbour's block-universe ideas, but made precise via algorithmic information theory. A striking consequence: in such a globally fixed configuration, correlations between hidden variables and measurement settings (superdeterminism) arise structurally, without any faster-than-light signaling. KT doesn't require this interpretation of quantum mechanics, but it permits it naturally.

The paper then maps KT against an unusually wide range of traditions — Aristotle's hylomorphism, Kant's transcendental idealism, Schopenhauer's will, William James's radical empiricism, Buddhist two-truths doctrine, Advaita Vedanta, Islamic occasionalism, and more. The comparisons are substantive, not decorative. Aristotle's "entelechy" (a thing's self-maintaining activity) becomes telehomeostasis — the agent's policy of maximizing its own future persistence. Kant's a priori forms of intuition become the architecture of the Modeling Engine. Schopenhauer's blind Will becomes the Objective Function. Islamic occasionalism's denial of local efficient causes maps onto KT's claim that causation is an emergent modeling tool, not a feature of the base. The paper is honest that the base's experiential nature is an axiom, not a derivation — but it argues the axiom does real work by dissolving the hard problem's "hard start" and by avoiding Benacerraf's access problem for mathematical knowledge.