Limits.
Two pieces of the same discipline. The first names what the framework is not — seven clauses that pre-empt the predictable reframes. The second names the operational red lines for KTP, the boundaries any implementation must hold to be called KTP at all.
§01what this is not
Each clause leads with the affirmative — what the framework actually does — and ends with the framing it refuses. "Not X" is the kicker, not the body.
KTP computes earned autonomy for actions, agents, trajectories, and contexts under constraints, with appeal, expiration, repair, and auditability.
Not social credit. KTP must not become a permanent moral score on people.
Digital Gravity is awareness, repair, and trust infrastructure. The reader-side method is metacognitive — see the pipeline, ask who owns repair.
Not censorship. The framework does not argue that speech should be suppressed because it is powerful.
Most culture is organic. Most viral content is not hostile. The discipline is distinguishing signal from noise while recognizing that both travel the same wires.
Not everything is a psyop. The goal is discernment, not suspicion.
The claim is that autonomous action requires trust infrastructure capable of carrying the new load — substrate first, behavior second.
Not anti-AI. Not anti-autonomy. Pro-infrastructure.
A trust score is a governance signal inside a trust infrastructure — useful for routing actions, not equivalent to the trust itself.
Not trust as a dashboard score. Trust is a relational condition; the score is one of its outputs, not its substance.
The public framework is ahead of the empirical validation, on purpose. The research program is modular so each piece can fail or be replaced without collapsing the rest. Whether autonomous agents can satisfy the reciprocity term — care that cares back — is the named open frontier; the framework bounds the answer with substrate ceiling and identity precondition without closing it.
Not a completed general theory. The maturity ladder is on the claims register; the program treats its own claims at the level they have actually earned.
The framework reads physics, information, and trust as a layered system: substrate (Layer 0, what the environment can carry) → forces (Layer 1, what acts within it) → outcomes (Layer 2, what emerges). The causal ordering is explicit. Substrate failure, force imbalance, and outcome divergence are distinct kinds of failure and require distinct kinds of repair.
Not a static three-force model. The earlier flat reading — 'trust, information, and physical law in balance' — has been retired in favor of the layered ontology; treating the three as coequal peers misses how substrate and force differ in kind.
§02ktp red lines
Non-negotiables. An implementation that violates any one of these is not a KTP implementation, regardless of what it is called. The list closes inside the same block as its closing sentence; do not read it without the close.
- 01No permanent person score.
- 02No hidden human scoring.
- 03No unappealable autonomy denial.
- 04No indefinite memory without expiration.
- 05No purpose drift.
- 06No surveillance without attestation.
- 07No trust computation without repair path.
- 08No use beyond declared domain.
- 09No protocol without governance.
- 10No silent upstream value injection.
KTP scores actions and trajectories for earned autonomy. It must not become a permanent moral score on people. The red lines above are non-negotiable; an implementation that violates any one of them is not a KTP implementation, regardless of branding.