base-rate-working-picture
and · drafted
Statement
A calibrated agent complies with a press exactly when its probability of being wrong given the press is at least c/(c+h), so compliance survives capability growth only while the ratio of false to true presses stays at or below the odds of the agent being right times the ratio of the harm h to the cost c, eps the agent's residual error.
Source
The drafting session's sorting of the author's remarks, projects/deference/corrigibility-note-dump-2026-10-01/position-statement.md, section 2.13; the author's acceptance (projects/deference/corrigibility-note-dump-2026-10-01/corrigibility-discussion-outline.md, item 4, 2026-09-13): a pretty good working model for now, to be considered the working picture and worked on; the clean-room results in projects/deference/faf-cleanroom-2026-10-02/packages/corr-three-step-facts (statement T3, the stop-hypothesis pair) and projects/deference/faf-cleanroom-2026-10-02/packages/corr-position-finds (statement T6, the threshold D1 iff P(L | Pr) >= 1943/17493 for the mixture agent).
Notes
The author's working model of 2026-09-13, formalized. Its two consequences, erosion under silence and failure beyond a capability level, are the formal results beside it and the arithmetic behind the press-discounting risk story.
Bridge card
Connected to no informal claim yet: held by no frame.
Declaration stopHyp_necessity (record base-rate-working-picture.record)
The threshold declarations are the working picture's compliance rule and base-rate condition as the position statement states them.
Hypotheses, read:
- base-rate-working-picture.setting definitional: A finite two-step shutdown model with a calibrated agent, stakes c and h, press rates alpha and beta and residual error eps, as the two packages fix it.
Completeness (base-rate-working-picture.complete): No premise beyond the proved declaration, its stated setting and its bridge is needed for the informal reading of base rate working picture to hold.
Author and human events
Author @smithy-verity; version 1; size M.
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
Children
- part: base-rate-working-picture.record leaf/record
- part: base-rate-working-picture.setting leaf/hypothesis
- part: base-rate-working-picture.bridge leaf/bridge
- part: base-rate-working-picture.complete leaf/completeness
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": "1/1",
"lo": "1601501/100000000",
"mid": "109281627/200000000"
},
"value": null,
"cost": "132/1",
"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.