Site Info Center

Writing Standard

• Evidence before abstraction. Every Node, relationship, or structural judgment must be grounded in observable evidence, real-world systems, documented capabilities, or clearly identified constraints.
• Structure before trend. Entries must explain why a development changes the Map, its canonical Primary Map Layer and terminal Primary Map Branch where applicable, and how it connects to other Layers. Isolated news summaries or generic technology commentary do not qualify.
• Cross-layer traceability. Important claims should be traceable across the relevant canonical Layers of the AI Civilization Map rather than through an informal or substitute layer list.
• Conditional, not deterministic. The map may identify structural pressure, directional advantage, or narrowing options, but it must distinguish observed facts from analytical inference and must not present uncertain trajectories as inevitable.
• Real-world anchoring. Companies, products, institutions, infrastructure, capacity, costs, deployment limits, and measurable operating conditions should be used whenever they clarify how a structural relationship works in practice.
• Defined conceptual boundaries. K Robot Lab does not expand through abstract philosophy, generic humanism, psychology, sociology, or speculative futurism unless those subjects are tied directly to observable systems and documented structural change.
• No advocacy or investment recommendation. Inclusion on the map does not imply endorsement, political support, commercial quality, future performance, or a recommendation to buy, sell, or support any entity.

Map Governance

Human reading order: Governance is defined from the largest structural object to the smallest: Layer → Canonical Branch → Observational Branch → Node → Relationship. Cross-object review and history rules follow only after those five governance levels are defined.

1. Layer — Architecture

• Layer semantics. A Layer is a canonical structural domain and civilization function. Layer numbers provide stable identifiers and reading order within a release; they do not establish causal priority, temporal sequence, vertical rank, maturity, or importance. The conceptual labels such as Foundation, Machine Substrate, Coordination Fabric, and Cognitive Layer describe functional roles rather than positions in a stack.
• Layer roles and ecological model. In v1.2, Layers 1–8 represent interdependent structural domains that share resources and continually enable, constrain, compete with, or redirect one another. Layer 9 represents the human social system of consequence, participation, legitimacy, and feedback; it is not a final technical stage. For human understanding, the current architecture may be visualized as eight structural trees in one shared ecosystem, with Branch Paths as branches, Nodes as leaves, and humanity living within and reshaping the forest. The official information hierarchy remains Layer → Branch Path → Node within each Layer; it does not describe causal flow between numbered Layers.
• Method architecture adaptability. The AI Civilization Map Method follows the Map’s current canonical architecture rather than imposing a permanent Layer count. Under v1.2 it operates across nine Layers; if a future major Map release changes the number or boundaries of canonical Layers, the Method applies its same evidence-first procedure to that revised full system. Cross-layer conversion is diagnostic and may operate in multiple directions; it is not a required bottom-to-top pipeline. This clarification does not change Method v1.0.
• Layer architecture and version policy. A material Layer Architecture Change — adding, removing, merging, splitting, or materially redefining a civilization-scale Layer — requires a Map version change because it changes the architecture to which the Method is applied. A pure wording or label clarification that leaves the Layer function intact is not automatically an architecture change. Branch Structure Changes, Branch Observation States, Node changes, and Relationship connectivity revisions do not by themselves require a Map version change, although an explicit governance or Method release may still be versioned by deliberate decision. The current five-level governance clarification does not change AI Civilization Map v1.2 or Method v1.0.

2. Canonical Branch — Topology

• Branch hierarchy. Canonical Branches form hierarchical structural paths within a Layer. Every formal canonical Branch object is part of topology whether terminal or intermediate. A canonical Branch must represent a durable, repeatable structural category and should not be created solely to accommodate one publication or Node. Descriptive path wording alone does not create a counted Branch object.
• Canonical Branch current states. A current Canonical Branch is either Populated or Reserved Empty. Populated means one or more active Nodes exist anywhere within that Branch path or its descendants. Reserved Empty means the structural category remains canonical while zero active Nodes currently qualify. Both states remain included in the Canonical Branch count, and a Branch may move between Populated and Reserved Empty without changing that count.
• Canonical Branch Structure Changes. Canonical Branch changes use Added, Deepened, Split, Merged, Renamed, Reorganized, and Retired. Branches do not use the Node classifications Promoted, Demoted, or Moved: a Branch changing parentage or structural position is a topology Reorganized event. Inserting a new intermediate Branch around an existing Node does not by itself constitute a Node move; a Node is Moved only when its canonical Primary Layer or terminal Primary Branch changes.
• Canonical Branch historical exit. A canonical Branch that no longer represents an independent, durable structural category is Retired. Retired Branches leave current topology and reduce the current Canonical Branch count, but their historical identity, prior path, and reason for exit remain preserved. When a Branch is absorbed or transformed through a merger, split, or reorganization, the more specific structural classification takes precedence over Retired.
• Branch identity and counting. Renaming or reorganization normally preserves the same Branch identity when the underlying structural function remains continuous. Split and merger events preserve predecessor and successor identities. The current Canonical Branch count includes every formal current canonical Branch object, including Reserved Empty and intermediate Branches, but excludes Observational, Closed, and Retired Branches. Creating a formal intermediate canonical Branch changes the count; merely adding descriptive path text does not.
• Terminal Branch. The Primary Map Branch shown in an article is the terminal Branch to which the Node is currently attached. It does not imply that no broader or intermediate Branches exist above it.

3. Observational Branch — Pre-topology Observation

• Observational Branch. An Observational Branch is a provisional pre-topology observation path used when a coherent structural problem space is visible but current evidence is not yet sufficient to admit it into canonical topology. It does not increase the Canonical Branch count, does not by itself authorize Node admission, and is excluded from canonical topology until formally admitted.
• Branch Observation State. Observational Branches use four observation-state actions: Opened when a provisional observation path begins; Updated when material new evidence changes the observed scope, mechanism, institutional trajectory, structural direction, or evidentiary basis while the Branch remains below canonical admission; Closed when continued observation is no longer warranted; and Reopened when a previously closed observation again receives material evidence. Updated records may carry an evidence direction such as Strengthened, Weakened, or Mixed. Closed Branches leave the current observation set but retain historical identity.
• Canonical admission. When later evidence demonstrates a durable, repeatable real-world structural function, an Observational Branch enters canonical topology through Branch Added, not Closed + Added. It then becomes either Populated or Reserved Empty according to its Node state. An observational stage is optional if evidence already meets the canonical threshold. Closing an observation is not Branch Retired because the Branch was never canonical.

4. Node — Structural Unit

• Cross-layer placement. The Primary Map Layer and Primary Map Branch identify the Node’s canonical location. Secondary and Supporting Map Layers describe dependencies, effects, constraints, or feedback relationships; they do not create additional primary locations.
• Admission threshold. A Node enters the active Map only when it adds explanatory value, has sufficient evidence, and can be assigned to a canonical Primary Map Layer and terminal Primary Map Branch, with cross-layer relationships identified where structurally material. Branch admission and Node admission remain separate thresholds.
• Research-role classification. Map-related research is not automatically an active Node. A publication may serve as Active Node Research, a Method Application / Map Test, or an Archived Reference. A month may contain many new articles and no new active Nodes.
• Archive distinction. An Archived Reference is research, method material, or a historical interpretation preserved for reference and does not imply that the publication was ever an active Node. An Archived Node is a formerly active Map unit that left the active topology but retains historical structural value, together with its prior Node identity and Branch Path. Archived Reference and Archived Node are therefore distinct governance states and must not be used interchangeably.
• Node identity, structural status, and knowledge state. Once admitted, a Node retains a historical Node identity. Its active status and structural position may change through Structural Node Changes, while the knowledge represented by the same active Node may also be materially updated without changing topology. Its governance record is not erased.
• Structural Node Changes. The Map uses six structural classifications for Nodes: Added, Removed, Archived, Promoted, Demoted, and Moved.
• Six Structural Node Change classifications. Reviews use six separate structural labels for Nodes:
  • Added. A new Node enters the active Map after sufficient evidence establishes a durable structural relationship, a canonical Primary Map Layer, and a terminal Primary Map Branch.
  • Removed. A Node leaves the active Map because the original structural judgment is no longer valid, useful, distinct, or sufficiently supported. Its prior admission and removal reason remain preserved in the Activity Log.
  • Archived. A Node leaves the active Map but retains historical structural value. Its Node identity and prior Branch Path remain preserved as part of the Map’s development history.
  • Promoted. An existing Node gains greater systemic importance or moves from a supporting role into a more central position.
  • Demoted. An existing Node remains valid but becomes less central because its scarcity, influence, or explanatory value has declined.
  • Moved. An existing Node is reassigned to another Layer or Branch because new evidence provides a more accurate canonical position; movement does not by itself imply promotion or demotion.
• Knowledge Change — Updated. Knowledge Change is recorded separately from Structural Node Changes and does not by itself alter a Node’s canonical Map placement, structural classification, or topology counts. Updated applies when an existing active Node remains in the same canonical Primary Map Layer and terminal Primary Map Branch and does not qualify as Added, Removed, Archived, Promoted, Demoted, or Moved, but its underlying evidence, scope, definition, or analytical content has been materially revised. Editorial, formatting, SEO, metadata, source-formatting, or other cosmetic changes alone do not qualify.
• Canonical Node Content Rule. An active Node article must represent the latest internally consistent structural interpretation supported by current evidence. When a Node qualifies as Updated, material new evidence should be integrated into the article’s existing reasoning, definitions, and conclusions as needed so the page remains a coherent current-state representation. Superseded claims should be revised or removed rather than preserved as competing narratives solely to retain edit history.
• Historical Change Record Rule. The Node article preserves the current canonical understanding; the AI Civilization Map Activity Log preserves the material revision history. Each Updated record should state when the Node was updated, what materially changed, why it changed, the evidence or scope added or revised, and whether the change affected structural status, canonical placement, or Map topology. Article metadata may identify the current Last Updated date, while the Activity Log preserves the sequence of material updates.
• Knowledge Change effective rule. The Updated classification was introduced on September 7, 2026 and is first applied in MR-2026-09. MR-2026-07 and MR-2026-08 retain the governance schema in effect when those reviews were recorded and are not retroactively normalized with an Updated zero state.

5. Relationship — Connectivity

• Relationship definition. A canonical Relationship is a directional, structurally material connection between active canonical Nodes, not merely a hyperlink, co-mention, thematic similarity, or correlation. Relationship connectivity is part of the Node’s current canonical interpretation rather than a separate topology count.
• Node-article surface rule. Every newly admitted Active Node article must include a standardized Related K Robot Research section near the end of the article, before the footer or attribution area. Each linked active Node in that section must carry an explicit Relationship semantic so both human and machine readers can understand why the connection matters.
• Relationship semantic rule. Canonical semantics are depends on, enables, constrains, converts, transmits, is downstream of, and forms a feedback loop with. Unless the label itself is reciprocal, the semantic is read from the current Node toward the linked Node: Current Node → relationship → Linked Node. A non-Node supporting or archived research link may still appear in Related K Robot Research, but it does not become a canonical Node-to-Node Relationship merely by being linked.
• Legacy and revision rule. Existing Node articles may already contain implicit connectivity through prose, cross-layer metadata, and internal links, so no bulk backfill is required. When a legacy Node receives a material revision, its Related K Robot Research block should be brought into the explicit semantic format where relevant.
• Activity Log boundary. Individual Relationship edges are not logged as separate Added, Updated, or Removed events in the Map Activity Log and do not receive independent monthly zero-state counts. If connectivity changes materially enough to revise the Node’s canonical interpretation, the affected Node is recorded once as Updated; the Activity Log may briefly note that connectivity was revised, while the detailed current relationships remain in the Node article.

Referential / Cascade Integrity

• Referential integrity / no-orphan rule. Every active Node must resolve to exactly one canonical Primary Map Layer and one valid terminal Primary Map Branch. Secondary and Supporting Layers describe dependencies or effects but do not create additional primary locations. A Node may not remain active under a Branch or Layer that no longer exists in current canonical topology.
• Cascade reconciliation rule. A parent structural change is not governance-complete until every affected descendant has been reconciled. A Layer Architecture Change must provide a migration mapping for all affected canonical Branches and Nodes. A Canonical Branch that is Retired, Split, Merged, or Reorganized must leave every affected Node on a valid successor path or formally classify that Node as Moved, Archived, or Removed as applicable.
• Persistent identity and alias continuity. A rename does not create a new Layer, Branch, or Node when the underlying structural object remains continuous. The canonical identity persists and prior names remain historical aliases or provenance references. Split and merger events preserve predecessor/successor lineage so later readers and machine systems can distinguish renaming from structural replacement.
• Relationship endpoint integrity. A canonical Relationship may connect only active canonical Nodes. When a Node is Removed or Archived, every inbound and outbound canonical Relationship involving that Node must be reviewed. Historical or supporting links may remain in article prose or Related K Robot Research, but they cease to be canonical active Node-to-Node Relationships unless both endpoints are active again under current governance.
• Downstream connectivity reconciliation. Relationship edges are not logged individually. If a Node exit or structural move forces a material reinterpretation of a surviving Node’s connectivity, that surviving Node is recorded once as Updated; otherwise the relationship cleanup remains part of Node-article maintenance and does not create a separate Activity Log event.
• Evidence-only, no-scoring rule. Formal Map states and changes are triggered by observable evidence of what has actually occurred. The Map does not assign confidence scores, importance scores, maturity percentages, relationship-strength ratings, rankings, probability ratings, or other subjective scoring systems to possible futures. Structural labels such as Promoted or Demoted must describe an evidence-supported change in real structural role or centrality, not a subjective preference score. Evidence-direction annotations such as Strengthened, Weakened, or Mixed summarize how new observed evidence affects an existing observation; they are not probabilities or forecasts.
• Governance completion invariant. No Layer change may leave unresolved Branches. No Branch change may leave unresolved Nodes. No Node exit may leave unresolved canonical Relationships. Until the required descendant and endpoint reconciliation is complete, the higher-level structural change is not considered fully reconciled in the canonical Map.

Cross-Object Review and Change Control

• Contradiction audit. New or revised Nodes must be checked against existing canonical Layer assignments, Branch definitions, structural functions, and cross-layer relationships. Apparent conflicts must be reconciled or explicitly scoped by time, geography, or system boundary before they enter the canonical Map.
• Revision over preservation. Existing Layers, Branches, Nodes, and Relationships must be revised when new evidence changes their structural meaning. Earlier interpretations are not preserved merely for narrative consistency.
• Classification precedence and no double counting. For the same governance object, a structural classification takes precedence over a state or knowledge update when the same evidence would otherwise trigger both. A Node Moved is not also counted as Node Updated; a Branch Merged or Split is not also Retired; an Observational Branch admitted to topology is Branch Added rather than Closed + Added. This rule prevents duplicate accounting of one object. Relationship connectivity is not counted as a separate Activity Log object: if a connectivity revision materially changes a Node without changing its structural classification or placement, the Node is recorded once as Updated.
• Monthly review. Each calendar month receives one canonical Map Review covering Layer Architecture Changes, Canonical Branch Structure Changes and current states, Branch Observation State changes, Structural Node Changes, non-structural Node Knowledge Changes, Archived References, and evidence that changes structural importance. Relationship connectivity is reviewed at the Node-article level rather than as a separate Activity Log stream; when a material connectivity revision changes a Node’s current canonical interpretation, that Node is recorded once as Updated. The review uses an explicit cutoff date and does not need to occur on the final calendar day.
• Review identity and cutoff. Each canonical monthly review receives a persistent identifier in the form MR-YYYY-MM and records its exact Review Cutoff date. The cutoff defines the topology state represented by that review and prevents a monthly label from being mistaken for an unspecified month-end snapshot.
• Baseline rule. A Node already present in the reference topology at the beginning of a review interval must not be classified as Added in that review. The first canonical monthly review may therefore include a Baseline Reconciliation section identifying Nodes already present in the t0 topology, while only topology changes occurring after that baseline are counted as monthly additions, removals, moves, promotions, demotions, or archival actions. For the first public AI Civilization Map, v1.0 was first published on July 15, 2026, while the canonical t0 topology was fixed on July 17, 2026; MR-2026-07 reconciles that t0 baseline to the July 31 review cutoff.
• Explicit zero-state rule. A canonical review must record confirmed zero changes rather than leave categories undocumented. All six Structural Node Change classifications — Added, Removed, Archived, Promoted, Demoted, and Moved — should be shown even when zero. Beginning with MR-2026-09, the separate Node Knowledge Change Updated, Canonical Branch Structure Changes, and Branch Observation States Opened / Updated / Closed / Reopened should also record zero states when none occurred. Archived References remain separate because they are research-role records rather than topology objects. Relationship connectivity has no independent zero-state category because individual edges are maintained in Node articles rather than counted in the Activity Log.
• Topology continuity record. Each review should preserve cutoff-state counts for Layers, Canonical Branches, active Nodes, and Reserved Empty Branches, together with the delta from the preceding canonical review when available. Reserved Empty is a subset of Canonical Branches and is not added again to the total. Populated + Reserved Empty therefore equals the current Canonical Branch total. Observational Branches are tracked separately and excluded from canonical topology; Closed and Retired Branches remain historical and are excluded from current counts. Relationship connectivity has no independent topology count; its canonical current state resides in the connected Node articles.
• Transparent change history. The Map page should present a concise latest-review summary; when Knowledge Changes are present, they should be shown separately from Structural Node Changes. The AI Civilization Map Activity Log should preserve the Review ID, Review Cutoff, explicit zero states, previous and new positions where applicable, evidence, material or zero canonical Branch Structure Changes, Branch Observation State changes, canonical topology counts and deltas, and the reason for each recorded Node or Branch event. Individual Relationship edges remain in Node articles and are not enumerated in the Activity Log. Historical monthly entries should use Removed from Active Map rather than the time-relative phrase Recently Removed.

K Robot Lab must remain revisable. Consistency matters, but fidelity to observable conditions takes priority over protecting an earlier framework, preserving a favored narrative, or forcing new evidence into an outdated structure.