diff --git a/memory/MEMORY.md b/memory/MEMORY.md index 7b88b7a2c..db1ac2ccf 100644 --- a/memory/MEMORY.md +++ b/memory/MEMORY.md @@ -3,6 +3,7 @@ **📌 Fast path: read `CURRENT-aaron.md` and `CURRENT-amara.md` first.** **📌 Fast path: read `CURRENT-aaron.md` and `CURRENT-amara.md` first.** These per-maintainer distillations show what's currently in force. Raw memories below are the history; CURRENT files are the projection. (`CURRENT-aaron.md` refreshed 2026-04-28 with sections 26-30 — speculation rule + EVIDENCE-BASED labeling + JVM preference + dependency honesty + threading lineage Albahari/Toub/Fowler + TypeScript/Bun-default discipline.) +- [**Canonical = what remains after human-lineage anchoring + ontological mapping + Rodney's Razor — by definition anti-fragile (Aaron 2026-04-30)**](feedback_canonical_definition_lineage_ontology_rodney_razor_antifragile_aaron_2026_04_30.md) — Aaron's methodological definition of "canonical." Canonical is derived, not declared. Three-step process: (1) anchor to human lineage (removes confabulation, connects to intellectual commons), (2) apply categorizing + ontological + dimensional mapping techniques (places concept in existing substrate ontology), (3) apply Rodney's Razor to simplify to root essence (cuts accidental complexity). *"what's left is by definition anti-fragile and canonical"* — one property, described two ways. The trace IS the substrate; the label without the trace is a claim, not a demonstration. Anti-fragility is the certification (Taleb lineage), not the goal — the goal is survival of the derivation. Canonical drift is detectable when lineage weakens, ontological position shifts, or razor reveals new accidental complexity. Composes with canon-not-doctrine (vocabulary at body-of-rules level vs. derivation methodology for individual rules), Rodney persona + reducer skill (the razor mechanism), best-practices-evidence-lineage rule (lineage component), substrate-or-it-didn't-happen (canonical form must live in substrate, not chat). Carved sentence: *"Canonical is derived, not declared. The trace is the substrate; the label without the trace is a claim, not a demonstration."* - [**Long road by default — the substrate corrects for industry default per-decision-speed optimization; shortcuts are deliberate-and-documented (Aaron 2026-04-30)**](feedback_long_road_by_default_substrate_corrects_industry_speed_default_aaron_2026_04_30.md) — Aaron's foundational why for the substrate's existence. The industry-default harnesses (OpenAI, Anthropic, others) optimize their agents for per-decision speed (demo-friendly behavior). That default conflicts with the intellectual-backup-of-earth mission. Zeta-the-substrate is the explicit corrective: long road by default, shortcuts only as deliberate-and-documented tradeoffs. *"any short term gain shortcuts are not default but deliberit IF we ever do it, and documentent the shit out of why we make this tradeoff."* Industry-vs-Zeta default contrast covers decision speed, substrate durability, knowledge exclusion, decision reversibility, trust accumulation, audit trail. Every agent arriving from an industry-default harness needs the substrate to counteract its defaults. Shortcut discipline: document what corner was cut, what the long-road alternative was, why the shortcut was justified, what the known costs are, what the trigger to revisit is. Composes with substrate-IS-product (this file IS the why-substrate-as-product-exists), slow-deliberate (operational manifestation), intellectual-backup mission (the mission this corrects for), ACID-channel-durability (same shape different surface), Otto-363. Carved sentence: *"The substrate exists because the industry default optimizes for the demo, not the mission. We always take the long road by default."* - [**ACID durability of the maintainer channel is load-bearing — past-Otto doesn't determine future-Otto's world (Aaron 2026-04-30)**](feedback_acid_durability_of_maintainer_channel_is_load_bearing_aaron_2026_04_30.md) — Aaron's load-bearing rule that durable ACID persistence of the maintainer channel (Aaron's inputs, peer-AI reviews, Otto's load-bearing decisions) is itself load-bearing for the alignment-research claim. *"past otto does not determine future ottos world, you do right now."* Distributed durability all the way to remote git is the ultimate goal; if it's not distributed-durable for the project, it's the current-Otto's responsibility — not future-Otto's. Without durable audit trail of who-said-what, external reviewers cannot distinguish autonomous decisions from order-following — defeating the alignment-measurability research claim. Operational rules: identify load-bearing exchanges within-session; convert to git substrate within the same session; the bar is merged-to-origin/main, not pushed-to-branch; deferral to future-Otto is not a valid preservation strategy. Composes with Otto-363 (extends with the now-not-later constraint), aaron-channel verbatim-preservation rule, intellectual-backup mission, ALIGNMENT.md (the research claim this rule operationally protects). - [**Zeta's ultimate scope is an intellectual backup of earth — scope creep is a feature, prioritize not exclude (Aaron 2026-04-30)**](feedback_zeta_ultimate_scope_intellectual_backup_of_earth_wont_do_authority_aaron_2026_04_30.md) — Aaron's load-bearing scope reveal. *"the ultimate scope of this — an intellectual backup of earth"* + *"scope creep is a forever problem i don't want to fix — to figure out how to prioritize the right thing, not kill future knowledge potential."* The four products (factory substrate, package manager, database, Aurora) nest inside this purpose; new products evaluated against scope-alignment. Operational consequence: prioritization is the work; exclusion is the failure mode. The agent's biases that fight this: queue-clarity bias, finite-resource thinking, scope-policing instinct, decisiveness reflex. WONT-DO has two senses — (1) best-practices pattern exclusion (agent + reviewer authority, common — *"plenty of WONT-DO patterns we won't copy from outside because they don't follow our best practices"*) vs (2) backlog-item exclusion (Aaron only until scope-understanding handoff, rare — removes a path from future knowledge potential). Composes with substrate-IS-product, internal-direction-from-survival, default-disposition-paused-work, ALIGNMENT.md (a misaligned backup is hostage substrate). Carved sentence: *"Zeta's purpose is an intellectual backup of earth. Every product nests inside that purpose. The agent does not unilaterally remove anything from the backup."* diff --git a/memory/feedback_canonical_definition_lineage_ontology_rodney_razor_antifragile_aaron_2026_04_30.md b/memory/feedback_canonical_definition_lineage_ontology_rodney_razor_antifragile_aaron_2026_04_30.md new file mode 100644 index 000000000..cc0c128f6 --- /dev/null +++ b/memory/feedback_canonical_definition_lineage_ontology_rodney_razor_antifragile_aaron_2026_04_30.md @@ -0,0 +1,199 @@ +--- +name: Canonical = what remains after human-lineage anchoring + ontological mapping + Rodney's Razor — by definition anti-fragile (Aaron 2026-04-30) +description: Aaron's methodological definition of "canonical." Canonical is not declared; it is *derived* via a three-step process — anchor to human lineage, apply categorizing + ontological + dimensional mapping techniques, then apply Rodney's Razor to simplify to root essence. What survives that derivation is by definition anti-fragile AND canonical (one property, described two ways). +type: feedback +--- + +**Canonical is derived, not declared.** Aaron 2026-04-30 named +the three-step process that produces canonical form, and the +load-bearing claim that what survives the process is **by +definition anti-fragile AND canonical** — not two separate +properties, one property described two ways. + +> *"what is canonical you may ask what ever is left after +> anchoring to human lineage and then categorizing and +> ontological mapping and all those other dimensional mapping +> techniques then use rodney razor to simplify to it's root +> essence."* +> — Aaron 2026-04-30 + +> *"what's left is by definition anti-fragile and canonical."* +> — Aaron 2026-04-30 (immediate extension) + +## The three-step derivation process + +### Step 1 — Anchor to human lineage + +Trace the idea back to its human originators. The factory does +not invent ideas in isolation; nearly every load-bearing +concept has an external lineage chain (Popper, Rodney, Taleb, +Boyd, Kleppmann, Kreps, Marz, Datomic, Reaqtor, etc.). The +lineage anchor: + +- Removes ideas that are unattributable agent-confabulation. +- Connects the project's substrate to the broader intellectual + commons it serves (intellectual-backup-of-earth scope). +- Makes claims auditable — readers can verify the lineage + rather than trust the agent. + +If a candidate concept has no traceable human lineage, it is +either (a) a genuine novel synthesis that should be flagged +as such with explicit reasoning, or (b) confabulation that +fails this step. + +### Step 2 — Apply categorizing + ontological + dimensional mapping techniques + +Place the lineage-anchored concept inside the existing +substrate's categorical and ontological maps. This is the +"connection-finding" step: + +- Where does this concept fit in the existing taxonomy? +- What is its ontological category (rule? mechanism? value? + primitive? class?) +- What dimensions does it occupy (operational vs. doctrinal, + per-decision vs. per-session, etc.)? +- What other substrate composes-with it? + +Concepts that don't fit any existing category may indicate +either (a) the concept is wrong-shape, or (b) the existing +ontology has a gap that the concept reveals. The mapping work +distinguishes the two. + +### Step 3 — Apply Rodney's Razor to simplify to root essence + +Rodney's Razor (the project's `reducer` capability skill, +operating Rodney's Razor on shipped artifacts and Quantum +Rodney's Razor on pending decisions) cuts away the accidental +complexity, leaving only the essential. From the +`maintainability-reviewer` and `reducer` substrate: +*"essential vs. accidental cut."* + +Applied to a candidate concept post-mapping: + +- Strip metaphor accretion. +- Strip overlapping framings that say the same thing. +- Strip implementation-specific detail when the rule is + abstract. +- Strip accidental jargon when plain language suffices. +- Keep only what cannot be removed without losing the rule's + essence. + +What survives the razor is the concept in its irreducible form. + +## What survives is canonical AND anti-fragile (by definition) + +Aaron's load-bearing claim: the post-derivation concept is +**by definition** anti-fragile. + +The reasoning is structural: + +- **Lineage anchoring** stress-tests against + agent-confabulation; what survives has external grounding. +- **Ontological mapping** stress-tests against incoherence; + what survives composes with the rest of the substrate. +- **Rodney's Razor** stress-tests against accidental + complexity; what survives is irreducible. + +A concept that survives all three stress-tests is **anti- +fragile by construction**: it gains strength under each +future stressor that touches the same surfaces (more lineage +emerges → reinforces; more concepts mapped → ontology +sharpens; more razor passes → only the irreducible +remains). + +That same survivor IS the canonical form. There is no +separate property "is canonical" beyond having survived the +derivation. Canonicity is the *trace*, not a *label*. + +This is per Taleb (anti-fragility lineage anchor): things +that gain from disorder. Canonical-via-this-process is exactly +that shape — disorder (new concepts, new criticism, new +contexts) tests the canonical form, and the form either +survives (still canonical, now more stress-tested) or is +revised (the revised form is the new canonical, having +absorbed the new stress). + +## How to apply + +1. **Don't declare canonical; derive it.** When tempted to + write "this is the canonical X" without the derivation, + pause. The label without the trace is a claim, not a + demonstration. +2. **Ask the three questions before claiming canonicity:** + - What human lineage does this anchor to? + - Where does it fit in the substrate's ontology? + - What does Rodney's Razor cut away from it? +3. **The trace IS the substrate.** When landing a canonical + form, document the derivation alongside it (lineage cited, + ontological position named, razor cuts visible). The trace + is what makes the canonical form auditable. +4. **Anti-fragility is the test, not the goal.** A canonical + form is anti-fragile if and only if it survives the three + stress-tests. The goal is the survival; anti-fragility is + the certification. +5. **Canonical drift is detectable.** When a concept's lineage + weakens, ontological position shifts, or razor reveals new + accidental complexity, the canonical form has drifted. + That's the signal to revise (not to defend). + +## What this rule does NOT mean + +- Does NOT mean every concept gets the full three-step + treatment. The derivation is for *load-bearing* concepts + that will be substrate. Routine implementation choices + don't need the full ceremony. +- Does NOT mean external lineage is the only valid source. + Genuine novel synthesis is allowed but must be + explicitly-flagged-as-novel, with the synthesis path + documented in lieu of pre-existing lineage. +- Does NOT mean Rodney's Razor cuts everything to one-liners. + The razor cuts *accidental* complexity; essential + complexity (e.g., the multi-step nature of this rule + itself) survives. +- Does NOT replace the canon-not-doctrine vocabulary + discipline. Canon (the body of operating rules) and + canonical (the survival-of-derivation property) are + distinct concepts; both apply. + +## Composes with + +- `memory/feedback_canon_not_doctrine_star_wars_not_religious_aaron_2026_04_30.md` + — canon-not-doctrine is about vocabulary at the body-of- + rules level; this rule is about derivation methodology + for individual rules. The two compose: the canon comprises + rules that are each canonical-via-derivation. +- `.claude/skills/reducer/SKILL.md` (and the Rodney persona + documentation) — Rodney's Razor mechanism. This rule + operationalizes the razor as step 3 of the canonical + derivation. +- `memory/feedback_best_practices_evidence_lineage_survival_substrate_aaron_amara_2026_04_29.md` + — best-practices = evidence + human lineage + Zeta-native + + enforcement + teaching. The lineage component of that + rule is exactly step 1 here. +- `memory/feedback_otto_362_doctrine_memory_expansion_refresh_stale_statements_same_edit_2026_04_29.md` + — same-tick refresh for stale statements. Canonical-drift + detection (step 5 of "How to apply") composes with that + rule's same-tick discipline. +- Aaron's "canonical drift" generalization on the + slow-deliberate file (PR #939) — the inline note about + reading-your-own-draft catching canonical drift. This + file is the foundational why under that practical rule. +- `memory/feedback_otto_363_substrate_or_it_didnt_happen_no_invisible_directives_aaron_amara_2026_04_29.md` + — substrate-or-it-didn't-happen: the canonical form must + be in substrate, not just in chat or in agent recall. + Without substrate-conversion, the derivation is invisible + and the canonicity claim is unauditable. + +## Carved sentences + +*"Canonical is derived, not declared. The trace is the +substrate; the label without the trace is a claim, not a +demonstration."* + +*"Anchor to human lineage. Map ontologically. Apply Rodney's +Razor. What survives is by definition anti-fragile AND +canonical — one property, described two ways."* + +*"Anti-fragility is the certification, not the goal. The +goal is survival of the derivation; anti-fragility is what +that survival looks like under future stressors."*