malign-hypothesis-window
leaf/step · drafted
Statement
Malign hypotheses can end or corrupt the training process before the asymptotic elimination the guarantees rely on applies.
Source
projects/deference/note-dump-2026-08-11/notes/li-deference.md, section 0.3: "Malign hypotheses will be eliminated by the training process eventually, if the process continues indefinitely, but realistically, malign hypotheses could cause catastrophic problems which end or corrupt the training process before that point."
Notes
A risk story in the maintainer's words, with role risk-story in this sub-frame; its outcome class is unplaced and the corrigibility branch may place it. It is the reason the finite-time crux matters.
Author and human events
Author @smithy-verity; version 1; size S.
No human event has touched this node.
Edges
Above
- part of formalism-open-questions-0-3
- root (risk-story) of formalism-open-questions-0-3
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": {
"J": 0,
"mid": "7/10",
"range": [
"1/5",
"1/1"
],
"width": "1/1"
},
"draft": "ai",
"root": {
"status": "open",
"Pr": {
"hi": "1/1",
"lo": "1/5",
"mid": "7/10"
},
"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.