AgentSeeResearch Notebook
version 1.0.0 · created 2026-04-08 · updated 2026-04-08

P5: Gains Persist After Removal

predictionhypothesisoriginal
ClaimCapacity improvements are maintained after the system is removed; the system builds the human's own regulatory capacity rather than substituting for it.
This claim fails if
If capacity metrics return to baseline after system removal.

Capacity improvements are maintained after the system is removed. The system builds the human's own regulatory capacity rather than substituting for it.

This is also the consent/capacity circularity test -- if the system cannot pass it, it fails its own criteria.

Candidate mechanism

The coherence-agency bootstrap. If the bootstrap operates as hypothesized, the system's role is maintaining preconditions for a self-reinforcing cycle that builds the person's own regulatory capacity. Gains persist because the bootstrap itself -- not the system's direct interventions -- is what built durable capacity.

Tests

E4 (removal/persistence trial).