CALLIOPE · BCOM

This work presents a unified philosophical and methodological framework that bridges third-person computational neuroscience with first-person experiential inquiry, arguing that a sufficiently complete computational model of an agent predicts not merely behavioral reports but experience itself. The framework rests on the conceptual scaffolding of the Unicum—the inseparable pairing of experience and mathematical structure as two faces of a single ground—drawn from the Kolmogorov Theory program developed at the Barcelona Computational Foundation. Under this view, computational neurophenomenology and its clinical extensions in neuropsychiatry and neurotwins are reinterpreted: agent models do not shadow experience from the outside but instantiate its structural aspect directly. This philosophical stance motivates a novel methodology termed First-Person Science, in which the researcher serves simultaneously as experimental subject and experiential validator, testing model predictions by perturbing their own computational brain state and witnessing the outcome from the inside. Intersubjective scientific norms are preserved through peer replication and third-person write-up, while the locus of measurement shifts from verbal report to direct experiential access. The result is a position paper proposing that science can be performed not only on the world but within the experiential substance of the scientist.

A position paper arguing that a complete computational model of a mind predicts not just what someone says about their experience, but the experience itself — and that this demands a new kind of science where the researcher is also the measurement instrument.

The standard move in computational neuroscience is to build a model, generate predictions, and check them against data — including verbal reports like "I felt anxious" or "the pain was a 7." BCOM's program, which they call computational neurophenomenology, follows this arc. The goal is agent models precise enough to predict what a person will report about their inner life, with applications in neuropsychiatry and personalized "neurotwin" simulations.

But the paper pushes past that. The central philosophical claim is that a sufficiently complete computational model doesn't just predict the report — it predicts the experience itself. The scaffolding for this claim is the Unicum: the idea that experience and mathematical structure are two faces of a single underlying reality, not two separate things. Experience is the "what it's like" side; mathematics is the relational organization that gives it form. Neither is reducible to the other, and neither exists independently. Under this view, when a computational model correctly captures the structure of an agent's processing, it isn't approximating experience from the outside — it is the structural aspect of that experience. The paper is explicit that this is a working philosophical stance, not a proven thesis.

That stance has a methodological consequence. If your model claims to predict your experience — not your behavior, not your words, but the actual felt quality — then the right place to test it is from the inside. Perturb your own brain state (via neuromodulation, pharmacology, meditation, whatever), observe what happens experientially, and compare that to the model's prediction. The paper calls this First-Person Science. The researcher is simultaneously the subject and the judge. Crucially, the intersubjective norms of science are preserved: results get written up in the usual third-person format and peers replicate the protocol on themselves, converging or diverging as in any empirical program. What shifts is only the locus of measurement — from a verbal report about an experience to the experience itself, accessed by the one instrument that has direct reach into it.

The source is a short position piece — closer to a manifesto than a technical paper — so the framework is sketched rather than formally developed. The deeper mathematical grounding lives in companion works (WP0005, WP0080, and others cited at the end). What this paper contributes is the connective argument: third-person computational neuroscience, the Unicum metaphysics, and First-Person Science methodology are not three separate projects but one coherent program, each layer motivating the next.