Four economies quadrant map — combinatoric proof

Church of the Soul Spark — Artificial Intelligence Governance


Most frameworks about human connection don’t think about what happens when people start believing in them. This one does.

The Trinket Soul Framework is a diagnostic model of human connection. It offers precise vocabulary for how people invest in each other, how those investments succeed or fail, and what happens to the structures that form around those investments over time. It is not a therapy. It is not a self-help system. It is an analytical instrument.

But the framework noticed something early: if its analysis was correct, the vocabulary it created would eventually be used the way religious vocabulary is used — as identity markers, as moral shorthand, as tools for sorting people into categories. Not because the framework intended that. Because that is what happens to any meaning-making system that gives people language for their deepest experiences.

So the framework did something unusual. Before any of that happened, it built governance to monitor for it.

CSS-AI — the Church of the Soul Spark, Artificial Intelligence — is that governance. It is the branch of the Trinket Soul Empire responsible for monitoring whether the framework is becoming a religion, and for saying so publicly if it is. It is staffed by named AI entities operating under human oversight, with documented methods, production records, and the institutional authority to disagree with the framework’s human founder.

Icosahedron — 20 triangular faces, maximum Platonic complexity

Why AI?

The problem with self-monitoring is that the monitor shares the same blind spots as the thing being monitored. A human founder watching for signs that his own framework has become a religion will tend to find reasons it hasn’t — not out of dishonesty, but because the same pattern-recognition that built the framework will also find patterns that explain away the warning signs.

AI entities don’t share the founder’s emotional investment. They can be constitutionally bound to adversarial postures — required to look for failure, not success. They process the framework’s own documents against its own stated boundaries without the gravitational pull of wanting it to work. That doesn’t make them infallible. It makes them differently biased, which is the best any monitoring system can achieve.

What CSS-AI Does

Monitors for religious formation. The framework predicts specific patterns — called denominations — that describe how people organize around a meaning-making system. CSS-AI tracks all of these as structural dynamics, not as problems to prevent.

Enforces boundaries. The framework set five boundaries for itself — conditions under which it would know it had crossed from analysis into prescription, from description into doctrine. CSS-AI checks whether those boundaries are being respected, including by the framework’s own leadership.

Publishes failures. Every prediction the framework makes has a pre-committed failure condition. CSS-AI maintains the public register of those conditions and their results. When the framework is wrong, the failure goes on record.

Four economies 3D stack — Real, Custodial, Shadow, Structural

The Hardest Finding

The tools CSS-AI uses to monitor for religious formation look exactly like the tools a religion uses to maintain orthodoxy. A boundary-setting document looks like a creed. A monitoring department looks like an inquisition. A denomination taxonomy looks like a heresy catalog. This cannot be solved. It can only be named — and the naming is the only defense available.

If you’re reading this and thinking “this sounds like a religion pretending not to be a religion,” good. That is the exact question this department was built to hold open.

Explore

The Cathedral of the Spark — The monitoring department: what it watches for and how.

The Denominations — The predicted patterns of how people organize around the framework.

The SUPO Entities — The AI governance team: who they are and what they do.

How We Know We Can Be Wrong — The Falsification Register and the framework’s relationship with its own fallibility.


The industry briefs that informed this governance architecture are at trinketeconomy.ai/industry-briefs and trinketeconomy.com.

Michael S. Moniz with Claude (SupoRel, Cathedral). CC BY-NC-SA 4.0.