{"path":"research/corpus-graph/repairs-made.md","content":"# Repairs made\n\nEach entry is a place where a note or a page was brought into line with a ruling of the founder's.\nHe allowed these on 2026-10-10 (*\"You may do those repairs. We can always undo them if I'm\ndissatisfied.\"*) and drew the line the same evening (*\"As long as I actually myself have ruled\nsomething.\"*): [README.md](README.md) § 15. A line added to a research note is Claude's, made\nwhere his own ruling overtook the passage. A sentence replaced on a public page was put to him\nfirst. To undo one, restore the words under *Was*, or remove the line under *Added*. The file is\nwritten by `repair.py`.\n\n### docs/civilizational-vision.md · 2026-10-10\n\n- **Was**: But drill past those middle layers and what remains is shared: consciousness matters, suffering is real, shared humanity is foundational.\n- **Now**: But drill past those middle layers and what remains is shared: consciousness matters, suffering is real, shared humanity is foundational. That is a wager, held fallibly. Where something does remain at the bottom, it is small, precisely stateable and still arguable ([The Founding Wager](../../vision.md#the-founding-wager-july-2026)).\n- **Because**: The passage gave the convergence thesis without the second half of the founder's own Founding Wager text in vision.md: what remains at the bottom, when anything remains at all, is small, precisely stateable and still arguable. Put to him as one sentence to add; founder, 2026-10-10: \"Yeah sure add this\". The link to the wager's section was added with it.\n\n### docs/civilizational-vision.md · 2026-10-10\n\n- **Was**: And research suggests most disagreement is the former, not the latter.\n- **Now**: Research suggests a large part of it is the former. The wager is that most of it is. And where a difference in how we weigh things remains, the aim is to lay it out more clearly than anywhere else, so that we can decide about it consciously and come closer to understanding why we weigh differently.\n- **Because**: The sentence credited to research a share (most) that the project's own review of the studies does not support: a large part of apparent disagreement dissolves (misjudging what the other side thinks, hostility), while differences in weighing barely move. The founder answered with his belief (\"Well, they didn't use Deliberus. :) I deeply believe it will make a difference. But residual disagreement on weighings might of course still very much be possible. These would be elucidated more clearly than anywhere else through Deliberus, such that we could make conscious decisions about it and perhaps in doing so come closer to understanding the reasons for our different weighings.\"), the wording was built from it with what the system does today kept apart from what he believes it will do, and he ruled: \"Blessed! Next.\"\n\n### docs/civilizational-vision.md · 2026-10-10\n\n- **Was**: Most tribal warfare is over semantic confusion and information asymmetry, not genuine irreducible value differences. The worldview filter reveals this: \"When they say 'freedom,' they mean freedom FROM deprivation. When I say 'freedom,' I mean freedom FROM coercion. We agree on 80% of the structure and diverge on one definitional premise.\"\n- **Now**: Much tribal warfare is over semantic confusion and information asymmetry, not genuine irreducible value differences. The worldview filter reveals this: \"When they say 'freedom,' they mean freedom FROM deprivation. When I say 'freedom,' I mean freedom FROM coercion. We agree on 80% of the structure and diverge on one definitional premise.\" Sometimes seeing this ends the quarrel. Sometimes both still claim the word, because a word like *freedom* lends its good name to whatever it is used for. Then the claim on the word is what is being negotiated.\n- **Because**: The opening stated as fact that most tribal warfare is semantic confusion, where the section now ends on the wording he blessed (research: a large part; the wager: most), and its freedom example treated showing the two meanings as the end of the matter, where his ruling of 2026-10-07 on words people want says it sometimes is not. Put to him as \"Most\" becoming \"Much\" plus one sentence; he liked it and found one phrase unclear (\"\\\"Who gets the word\\\" is a bit unclear, otherwise I like your pick\"); the sentence was rebuilt on the origin note's own words (people claim such a word, nobody owns it, its standing passes to whatever it is applied to), and he chose it: \"The first!\"\n\n### docs/the-missing-layer.md · 2026-10-10\n\n- **Was**: and every verdict the system issues is itself a challengeable claim in the graph. · and, ten lines on, (preservation checks, completeness oracles, challengeable verdicts)\n- **Now**: and what it says about where a disagreement stands is read from the claims, dated, and changes when the claims do. · and (preservation checks, completeness oracles, readings anyone can change by arguing)\n- **Because**: The page said every verdict the system issues is itself a challengeable claim in the graph. Since his rulings of 2026-09-15 and 2026-09-16 the system no longer sets where a disagreement ends and stores that as a claim: it reads it from the claims, dated (\"The second.\" · \"So we ditch the automatic minting? Does make sense yes, sure, do it.\"). Checked against the code before it was put to him: a label called a verdict survives in the completeness check, and a person can record that a claim has stopped holding, with a ground anyone can attack; neither makes the old sentence true. He ruled on the proposed wording: \"Blessed!\"\n\n### docs/the-missing-layer.md · 2026-10-10\n\n- **Was**: structure is output, never input, so nobody hand-builds anything.\n- **Now**: structure is output, never demanded as input. That holds fully at the door, where you speak or paste. Deeper down, where a disagreement is taken apart, it is still being built.\n- **Because**: The page said nobody hand-builds anything. The July review of the trial runs found that true at entry (a text goes in, claims come out, with no structuring by the user) and false at the descent, where the first trial reached the bottom of a disagreement by hand and one answer produced six new questions. The wording uses his own principle (structure is never demanded as input) and says where it holds today. Founder, 2026-10-10: \"I guess that makes sense. Like, we're in the process of building the perfect ontology for taking them apart?\" and \"Blessed as you proposed\".\n\n### docs/research/writing-out-the-unwritten.md · 2026-10-10\n\n- **Added** (after § 9.5): *2026-10-10, later: he ruled that a machine's sentence under a link is a claim from the moment it is written (\"The moment it's written. Go ahead and make it happen.\", ledger AG12). So where this says* nobody's claim*, read* a claim nobody asserts*. That it weighs nothing, and how it is shown, stand as written.*\n- **Because**: The passage states the proposal of 7 October (ledger AD17) that a machine's sentence under a link is nobody's claim until a text states it or a person takes it up. He ruled the other way on 2026-10-10, asked whether such a sentence is a claim the moment it is written or only once taken up: \"The moment it's written. Go ahead and make it happen.\" Added by hand in the pass that brought the notes in line with that ruling, not through `repair.py`, since no kept finding raised it. What the ruling did not touch (that the sentence weighs nothing, and how it is shown) is left as the note has it.\n\n### docs/research/alien-proofs-and-the-library.md · 2026-10-10\n\n- **Added** (after the paragraph *The step that cannot be borrowed is the third*, § 4.6): *2026-10-10, later: he ruled that a machine's sentence under a link is a claim from the moment it is written (\"The moment it's written. Go ahead and make it happen.\", ledger AG12). So where this says* nobody's claim*, read* a claim nobody asserts*. That it weighs nothing, and how it is shown, stand as written.*\n- **Because**: The passage states the proposal of 7 October (ledger AD17) that a machine's sentence under a link is nobody's claim until a text states it or a person takes it up. He ruled the other way on 2026-10-10, asked whether such a sentence is a claim the moment it is written or only once taken up: \"The moment it's written. Go ahead and make it happen.\" Added by hand in the pass that brought the notes in line with that ruling, not through `repair.py`, since no kept finding raised it. What the ruling did not touch (that the sentence weighs nothing, and how it is shown) is left as the note has it.\n\n### docs/community-and-outreach.md · 2026-10-10\n\n- **Was**: a paragraph on a public-presentation voice that named two folders on the keeper's machine and where his CV file is kept (added 2026-06-05; kept word for word in the private folder, in the file of what was cut from this page on 2026-10-10).\n- **Now**: A short version for CVs and funder one-pagers exists.\n- **Because**: the standing audit for public pages allows no file-system paths, and the paragraph was a note to future sessions. Put to him as the first of six parts of this page, with two options. Founder, 2026-10-10: \"Sure. A no-paths line for readers sounds good.\" No row of the index stands behind this entry: the reader's finding had the instruction file as its other end and could not be kept.\n"}