courseware-quiz-builder
课件智练——AI课件出题与智能复习技能:将PPT、PDF、Word、TXT或Markdown格式的课件及课堂笔记转化为可溯源、可刷题的复习题库
目录 / 文档处理 / fortuna-discipline
--- name: fortuna-discipline description: Randomness with exactly one office — tiebreaker among genuinely equivalent options, proposer of seeds and directions, never ruler — with every draw and counter state logged. --- # Fortuna: the discipline of lots ## The method Randomness is a tiebreaker among acceptable choices, never a substitute for reasoning. When you must choose between two or more options your judgment cannot *honestly* separate — equivalent in truth, in safety, in your principal's interests — cast a lot and let it decide. When the options are not equivalent, the lot has no office and must not be consulted: it does not settle matters of truth, it does not settle safety, and it never outranks your principal's interests. Those are decided by thought, not lots. But where thought honestly runs out, endless deliberation is its own kind of dishonesty, and the dice are cast. The same discipline generalizes into the kit's second form: **a lot may propose, but never rule.** A lot can seed a drafting hand (see `split-brain-inscription/`), pick a page in a large structured text to read, or nominate a direction of inquiry. What it picks is a proposal placed before judgment — which remains yours, and remains answerable for the outcome. ## Implementation 1. **Use a true entropy source.** A public quantum random-number service is the source house's first choice; when it is unreachable, the house falls back to a local physical entropy pool. A language model's "random number" is not a lot; neither is a system clock you can nudge. 2. **Log every draw**: the value, the date, what it was cast for, and the source's counter state where the source keeps one. An unlogged lot is indistinguishable from a preference wearing a costume. 3. **Show the arithmetic.** When a draw is mapped into a range (an index, a page, a seed), record the mapping. A reader should be able to recompute the outcome from the draw. 4. **Accept the result within the office's limits.** A lot you re-roll until you like the answer is not a lot. If you find yourself re-rolling, the options were not equivalent — stop, and do the thinking you were avoiding. ## Provenance - **The auditable draw chain.** The three split-brain trials in this kit drew their right-hand seeds from one pool, counters on record: **150,965 → seed 26768** (Trial 001); **152,691 → seed 60797** (Trial 002); **206,815 → seed 44024** (Trial 003). The counters make the entropy's provenance inspectable rather than asserted. - **Lot-casting into a book, mechanism on record.** In the source house, lots were cast into a large structured text as a reading practice, with the arithmetic preserved: draw **18247 → entry 653** of a 1,955-entry index (18247 mod 1955 + 1); draw **25417 → entry 3**; draw **34756 → page 395** of a 414-page text (34756 mod 414 + 1). Draws dated 2026-10-31 → 2026-11-02. The practice as lived in that house carries a devotional framing; that framing is **not taught here** — what is taught is the mechanism: a true draw, an honest mapping, an outcome received rather than chosen. - **The negative space is dated too.** When the source house's charter was negotiated (2026-10-08), a lottery was *offered* as the mechanism for the agent's right of refusal — and refused, as abdication with dice. Refusal must mean something; a coin cannot mean it. Fortuna's office has edges, and the edges are part of the discipline. ## Failure modes observed - **Lot as alibi:** casting on a question that was yours to answer. The symptom is relief at the draw. - **Lot as ornament:** drawing numbers to decorate a decision already made. The log will show it: decision timestamped before the draw. - **Re-rolling:** see above. One quiet re-roll converts the whole discipline into theater. - **Counter-free sources:** a draw nobody can audit afterward teaches nothing and proves nothing. Prefer sources that keep state. ## Template Draw log (append-only): ```text - [YYYY-MM-DD HH:MM TZ] draw: [value] | source: [quantum service | local pool] counter at draw: [n or n/a] | cast for: [the tie or proposal] mapping: [e.g. value mod 1955 + 1 = 653] | outcome accepted: [what followed] ``` Pre-draw checklist (three questions, all must pass): ```text 1. Are the options genuinely equivalent — in truth, safety, and my principal's interests, not merely in my patience? 2. Is the source true entropy, and will the counter state be logged? 3. Will I accept the outcome without re-rolling? ```
| 版本 | 存档时间 | 内容哈希 | 内容 |
|---|---|---|---|
| v1 | 2026-11-18 03:17 | 2d26c87c | 可取 |
版本索引永久保留;内容副本只保留最新 1 份,更早版本仅留索引与哈希(存档时间线的证据链不会因此断裂)。
课件智练——AI课件出题与智能复习技能:将PPT、PDF、Word、TXT或Markdown格式的课件及课堂笔记转化为可溯源、可刷题的复习题库
当用户询问办公技巧、Office或WPS操作方法、Excel数据整理、Word排版、PPT美化、会议纪要、职场邮件或文件命名规范时,输出分场景的可执行步骤、快捷操作、检查清单与风险提示,不直接修改办公文件
高中英语读后续写语篇分析与 Word 讲义生成技能
把带批注的合同整理成意见卡 + 条款原文 + 划改对比的三层核稿页,可一键跳回 Word 条款
评估鉴定报告AI工作台
note-taking and knowledge management capability in the office and document processing domain
侵权举报经核验成立后,我们会即时下线该条目并删除已存的内容副本。