From the wave-0 evidence ledger
Recorded claims
Each claim states what the inspected source says, at the recorded location—bounded by its scope and graded by its confidence. Nothing here is a synthesis across works.
The model frames decision authority as a trade-off between an AI system's alignment with the principal's objective and a human agent's ability and incentive to acquire decision-relevant information.
Scope: Stylized authority-allocation mechanism, not measured organizational behavior.
Even when AI has strong decision quality, assigning authority to a human can be optimal when human authority preserves information-acquisition effort; therefore prediction accuracy alone does not determine an efficient accountability boundary.
Scope: Under the model's assumptions; requires empirical calibration before application.
Boundaries
Limitations & independence
Recorded at coding time, carried with the work forever. A claim without its limits is not evidence.
Recorded limitations
- Stylized principal-agent model
- Alignment and effort parameters are not directly estimated
- Predates modern generative AI and agent products
Source independence
Independent academic formal model; no empirical source dependence.