AIM Framework glossary
Early-Identification Principle
Canonical definition
Canonical layer. The principle that a developing conflict, and the Confabulation Cascade underlying it, is arrested most cheaply and with the fullest prevention of A- and I-deprivation while the cascade is still subcritical, below the Subcriticality Threshold (P_c) and before Bayesian Lock-In. As the cascade compounds toward lock-in, and the system advances through crisis-space past the Calcification Threshold, the cost of arrest rises and its preventive effect falls non-linearly, until intervention can address only aftermath. Early identification therefore requires substrate disaggregation, determining whether the genuine stake is A-, I-, or M-dominant, since the corrective pathway differs by substrate.
Explanation
A conflict, or a confabulation cascade more broadly, is easiest to stop before the shared narrative hardens. Early on only a few minds hold it, so a light touch can change them and no one is deprived of anything. The difficulty then rises with scale: as the narrative propagates, each new adherent becomes social proof for the next, so arresting it later means overturning a belief already held by many and self-reinforcing through that very social proof, not merely persuading a few. Past lock-in the narrative can no longer be seen as a narrative at all; correction becomes prohibitively costly, then impossible, leaving only cleanup. Acting early also means diagnosing whether the real stake is material (A), meaningful work and autonomy (I), or status (M), because the remedy differs by case.
Epistemic status
Framework-Specific Principle.
Derivation
Related terms
- Conflict-as-Failure-Mode Principle
- Mechanic's Obligation
- Physician-Heal-Thyself Principle
- Confabulation Cascade
- Subcriticality Threshold
- Bayesian Lock-In
- Crisis State Architecture
- Crisis State Variables
- Two-Question Diagnostic
- Knowledge Patch
- Epistemic Suspension
- Non-Malice Corollary
- Non-derogable essentials
- Observability Gradient