policy-level-total-trust

Statement

What is the policy-level analogue of Total Trust for an AI trained to act and trusted directly, rather than an estimator run through an argmax rule, and does the bounds-transfer and self-endorsement factorization exist without a selection to condition on?

Source

projects/deference/note-dump-2026-08-11/wiki/open-problems.md, Research direction: trust the policy, not the decision rule: "model an AI trained to act — a policy the humans can trust directly — rather than an estimator run through an argmax rule"; "what is the policy-level analogue of Total Trust, and does an analogue of the bounds-transfer/self-endorsement factorization of [total-trust-implies-value] exist without a selection to condition on? It is not known whether this is doable".

Notes

A question: Value is instrumental and no decision rule is optimal across embedded environments, so any argmax-shaped statement inherits a decision theory; the shape of a statement that does not is unknown. A shared claim of the bounded-agent programme, and the target of the policy-level route of the collapse's improvement claim.

Its evidence parent is the deck's scoped chain read as an implication: that chain is the bounds-transfer and self-endorsement factorization for argmax choice, whose policy-level analogue the question asks for. The action is a one-page statement attempt.

Author and human events

Author @smithy-verity; version 1; size M.

No human event has touched this node.

Edges

Above

• part of deference-toward-bounded-agents
• given of deck-collapse-statement
• root (shared-claim) of deference-toward-bounded-agents
• given deck-collapse-statement (as valid)