udt-over-bli-tiles
leaf/step · drafted
Statement
A UDT 1.0 over a BLI-structured prior with real-value coherence and universal instantiation satisfies a tiling theorem in which earlier agent-moments do not prefer to modify later ones.
Source
projects/deference/faf-cleanroom-2026-10-02/sources/references/bli/understanding-trust/overleaf-final-source/main.tex, section Other Rationality Constraints: "We conjecture that some version of these assumptions allows UDT 1.0 to be adapted to a computationally uncertain setting, enabling more realistic (if still woefully unrealistic) tiling results." and, of the sketched conditions: "these ideas are still quite preliminary".
projects/deference/faf-cleanroom-2026-10-02/packages/udt-bli-tiling/udt-bli-tiling-report.md: one-step UDT tiles under independence and fails without it; the sequential single-coin mugging gives a re-freezing dilemma; the Vingean form is a prose OPEN.
Notes
The shared tiling step of every route of the direct route, given the finite updateful-when-appropriate lemma's validity as evidence: an agent that behaves updatefully exactly when it has no reason not to is the kind of agent a tiling theorem can be stated for. Finite shadow only; the LI-level statement is the action.
Author and human events
Author @smithy-verity; version 1; size L.
No human event has touched this node.
Edges
Above
- part of direct-route-via-b1
- part of direct-route-via-b3
- part of direct-route-via-linked-b2
- part of direct-route-via-luv-coherence
- given of updateful-when-appropriate
- given updateful-when-appropriate (as valid)
Ranking (the lens, not the structure)
{
"status": "open",
"diagnosis": null,
"flags": [
"AI number",
"given updateful-when-appropriate@valid"
],
"V": {
"V_prior": null,
"display": "2",
"headline": 2,
"ref": 2,
"valid": 2,
"weaker": "both"
},
"belief": {
"J": 0,
"mid": "3/10",
"range": [
"0/1",
"4/5"
],
"width": "1/1"
},
"draft": "ai"
}Say something about it
Each verb opens a prefilled issue; a round files what lands as the ledger event it is.