Alignment research asks what an AI system should value. TEI asks a prior question – what epistemic structure does the system instantiate?
