basin-four-questions
leaf/question · drafted
Statement
For an AI participating in recursive self-improvement, the basin's distance metric, what the AI must be correct about, what it may be wrong about, and when errors self-correct over self-improvement have answers.
Source
The author's own notes, projects/deference/note-dump-2026-08-11/notes/li-deference.md, section 0: "In what ways might an AI (such as a modern LLM) be \"good enough\" to participate safely in RSI (eg, constitutional AI, deliberative alignment)? What's the appropriate distance metric (if any)? What does the AI need to be correct about, and what can it be wrong about? When are errors self-correcting over the course of self-improvement?"
Notes
The author's four questions, not taken up as stated by any file; the toy basin fixes one metric, which the source does not choose.
Author and human events
Author @smithy-verity; version 1; size S.
No human event has touched this node.
Edges
Above
- part of dynamic-conditions-for-going-well
- root (shared-claim) of dynamic-conditions-for-going-well
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": "17/40",
"range": [
"0/1",
"37/40"
],
"width": "1/1"
},
"draft": "ai",
"root": {
"status": "open",
"Pr": {
"hi": "37/40",
"lo": "0/1",
"mid": "17/40"
},
"value": null,
"cost": "inf",
"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.