above-quantifiable-guides-engineering
or · drafted
Statement
A formal theory of trust developed at the level of probability frames and Jeffrey–Bolker models guides AI safety engineering work.
Source
projects/deference/faf-cleanroom-2026-10-02/sources/unbounded-embedded/01-primary/summer-research-plan-2026--research-ideas-may-2026.md: "The interpretation that's relevant for us: humans should happily trust AI to make decisions on their behalf precisely when humans totally-trust AI beliefs."; "One big problem with Managing Misalignment is that the approach is not constructive at all"
Notes
Member of the level family. Two routes: through translation to a computability level (conditioned on aq-starting-place) and directly (conditioned on the task aq-engineering-note). The two quotations do not pull against each other: the second is the plan's reason for moving to the logical-induction level (constructiveness-favors-li).
Author and human events
Author @smithy-verity-opus55; version 1; size S.
No human event has touched this node.
Edges
Above
- part of stance-and-level
- root (shared-claim) of stance-and-level
Children
- part: aq-guides-via-translation leaf/step
- part: aq-guides-directly leaf/step
Ranking (the lens, not the structure)
{
"status": "open",
"diagnosis": null,
"flags": [
"AI number"
],
"V": {
"V_prior": null,
"display": "2",
"headline": 2,
"ref": 2,
"valid": 2,
"weaker": "both"
},
"belief": null,
"draft": "ai",
"root": {
"status": "open",
"Pr": {
"hi": "7311/8000",
"lo": "0/1",
"mid": "3383461/10000000"
},
"value": null,
"cost": "33/1",
"margin": "no established leaves"
}
}Say something about it
Each verb opens a prefilled issue; a round files what lands as the ledger event it is.