arg-thornley-completeness
leaf/step · drafted
Statement
Agents with complete, transitive, option-set-independent, patient preferences try to prevent or cause the pressing of the shutdown button, for an agent whose utility on the correction algebra depends on the corrections.
Source
Thornley 2023, abstract and three theorems; projects/deference/corrigibility-note-dump-2026-10-01/corrigibility-discussion-outline.md, item 5; the clean-room package projects/deference/faf-cleanroom-2026-10-02/packages/lit-shutdown-prefs proves the First, Second and Third Theorems (see thornley-three-theorems) and the 2025 chain's antecedent is empty for expected-sum-total within-length agents.
Notes
Outline item 5. Its given edge from the open Lean statement on POSL and ILPACS: if no relation with a strict preference satisfies both, Thornley's own 2025 escape closes and the argument stands stronger against every agent, this one included.
Author and human events
Author @smithy-verity; version 1; size M.
No human event has touched this node.
Edges
Above
- part of anti-naturality-arguments-as-stated
- given of posl-ilpacs-nontrivial-model
- given posl-ilpacs-nontrivial-model
Ranking (the lens, not the structure)
{
"status": "open",
"diagnosis": null,
"flags": [
"AI number",
"given posl-ilpacs-nontrivial-model"
],
"V": {
"V_prior": null,
"display": "2",
"headline": 2,
"ref": 2,
"valid": 2,
"weaker": "both"
},
"belief": {
"J": 0,
"mid": "9/20",
"range": [
"0/1",
"19/20"
],
"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.