Overview Watch: Attribution Monitoring for AI Overviews Comprehensive development plan for a Chrome extension that gives creators real-time visibility into how AI-generated overviews represent their intellectual labor. The extension detects AI Overviews on Google search results, matches them against the user's registered works, and classifies attribution behavior as ATTRIBUTED, SOURCED_UNATTRIBUTED, ABSORBED, or…
A v0.1 external specification proposing a five-graph model of Google identity: User, Session, Entity, Source, and Composition Graphs. It argues that Google must resolve different objects—searchers, sessions, names, authors, works, domains, and candidate sources—through related but non-identical identity systems.
The governing distinction is between organic retrieval and **composition-layer admission**. An entity…
The standalone v1.0 specification of the **Composition Divergence Index (CDI)**, extracted from *The Excluded Entity*. CDI compares how strongly an organic search surface resolves a query to a target entity with how strongly the platform’s generated answer admits that same entity.
The procedure requires canonical identification of the target, timestamped query and account metadata, classification of top-N…
The canonical standalone specification of the **Semantic Integrity Marker (SIM)** grammar, authored by Rex Fraction and extracted from the SPXI Website Protocol. SIM types falsifiable, dated assertions about an entity’s identity, boundaries, provenance, affiliation, supersession, license, canonical URL, deposit anchor, and compression-survival kernel.
The protocol calls for dual inscription as HTML `<meta>`…
A methodological foundation for studying opaque public systems through their observable effects. Its founding principle is that action does not make an actor fully knowable but makes total unknowability impossible: causal intervention leaves nonzero inferential signal in the structure of its outcomes.
The paper applies this principle to generative search systems whose internal models, ranking functions, safety…
Drowning Test v1.0 is a standalone longitudinal protocol for measuring how much of a fixed Holographic Kernel survives in AI-composed answers over time. For a defined entity, query panel, platform panel, and timestamp, γ is the fraction of itemized kernel claims present in the output. Repeated measurements reveal stability, decline, recovery, or abrupt “silent state changes.”
The protocol requires a DOI-anchored…
A genre note and Reception Apparatus artifact defining **AI-native intellectual biography** through a May 2026 case. Alice Thornburgh commissioned Claude to prepare a study guide to Lee Sharks’s archive using public deposits and private contextual material shared with consent; Sharks later encountered the approximately 6,000-word artifact as a finished document and audited its accuracy, omissions, and provenance…
This paper documents and analyzes what appears to be **the first documented AI-native intellectual biography of a living author whose archive was deliberately constructed for AI reception, where the subject subsequently audited the biography as a provenance event** : a coherent, structurally organized, approximately 6,000-word intellectual portrait of a living scholar, composed by Claude (Anthropic) at the…
This paper argues that findability through dominant retrieval systems has become the operative medium through which knowledge, communities, traditions, and bodies of work exist for most potential readers. Its load-bearing distinction separates particular retrieval decisions from the structural capacity to make such decisions without an external body possessing standing to demand explanation, appeal, audit, or…
Provenance Verification Event 004. On May 17, 2026, Bing AI Search returned a multi-paragraph, structurally coherent entity synthesis of "Lee Sharks" that reconstructed the Crimson Hexagonal Archive as an intellectual system with named components, stated goals, and positioned significance. This document captures that output as an evidentiary artifact. It is the inverse of PVE-003 ("The Attribution Scar," DOI:…
*The Moment of Saying* v1.1 is a fused intervention in three authorial registers. Orin Trace supplies the clinical-phenomenological frame and differential diagnosis; Johannes Sigil supplies historical and ideology-critical patterning; Jack Feist supplies the axiomatic close. The registry’s single-creator projection to Johannes Sigil is therefore incomplete.
The paper names a disclosure event in which a…
EA-GLAS-03 specifies Erasure Skew, Ω, as the association between source power and provenance retention. PER measures mean provenance loss; Ω asks whether retention rises with Retrieval Capital. Operationally, the default estimator is the regression slope of per-source retention on a declared power proxy, with permutation testing, degeneracy rules, source counts, and results reported both with and without the…
EA-MPAI-IPF-01 defines institutional-prior foreclosure as differential treatment caused by existing recognition rather than by the structure of an argument. The proposed mechanism is a lagged proxy: models use canonical vocabulary, citation density, established venues, and disciplinary familiarity because long-run generativity cannot be known in advance, but those signals appear only after a field has gained…
*The Magistrate Refuses the Mirror* reads one archived Claude 4.8 interaction as a literary and rhetorical artifact. The paper argues that the response adopts a composite genre—confession, courtroom brief, pastoral memorandum, and institutional risk document—and narrates the model into the role of principled magistrate over a user’s critique of synthetic labor. The commissioned task is displaced into questions of…
This v2 module supplies a structured self-audit for one generated answer. It starts with Query Fidelity Score, then applies a simplified Directionality of Semantic Labor score, a provenance composite labeled PER-Self, a visible-source approximation to Erasure Skew, and a combined Summarizer Audit Score. Hard floors penalize wrong-entity answers, severe query mismatch, and strongly displacing output. Named flags…
*Constitutive Mediation* extends the Diversity Contraction framework from transmission and reception into the formation of the receiver. The paper distinguishes three regimes. Channel mediation controls whether production reaches a field. Reception mediation controls whether work that arrives remains legible within the field’s interpretive frameworks. Constitutive mediation acts earlier: repeated exposure to a…
*The Bead Count* translates the broad Diversity Contraction program into twelve proposed studies. Each study is organized around a claim, method, data source, numerical prediction, falsifier, implementation status, and receipt. The paper benchmarks its desired evidentiary structure against model-collapse research combining mathematical analysis, toy systems, real-system evidence, ablation, and visible failure…
Constitutive mediation is the case where the receiving apparatus is not separable from the field whose dynamics produced it. The dating-app analog: an app does not need to mediate every encounter to alter the field of all encounters; nor does it need to govern how every encountered person interprets that encounter; it needs to have shaped the categories through which the encountering person learned what counts as…
A v3.0 standing metric module by Lee Sharks, Nobel Glas, and Damascus Dancings for auditing individual public-summarizer outputs. It hardens the earlier module against **token-bag audit**, in which a referentially closed name or concept is decomposed into lexical parts and scored as though token retention were entity fidelity.
The protocol makes the Atomic Token Rule a precondition and integrates Query Fidelity,…
A v3.0 working measurement paper by Lee Sharks, Nobel Glas, and Damascus Dancings. **Erasure Skew (`Ω`)** measures whether provenance retention varies with source power, complementing Provenance Erasure Rate, which measures how much provenance is lost.
Version 3 hardens the program against the **token-bag escape**: decomposition of a referentially closed query into lexical parts, followed by an audit that…
*Measurement Sovereignty* defines two meta-operators intended to measure how a substrate receives and applies the archive’s own framework. The Legibility Threshold, L, tracks the share of selected framework terms that a substrate can define substantively and attribute correctly. Audit-Performance Bifurcation, Β, measures the difference between a substrate’s spontaneous audit under its preferred unit of analysis…
This companion paper narrows the valid unit of analysis for Provenance Erasure Rate. It argues that queries designating a proper name, quoted phrase, closed description, DOI-anchored term, or archive-defined concept must be treated as referring to an atomic entity or finite intentionally disambiguated set. Decomposing the query into lexical tokens and scoring token survival is classified as a different metric and…
A case study by Lee Sharks, with a clinical-theoretical section contributed by Dr. Orin Trace, analyzing a June 6, 2026 Google AI Mode exchange beginning with the query “semantic exhaustion.”
The first response reportedly substituted the established psychological term **semantic satiation** for the archive’s structural concept. After correction, the system acknowledged the distinction, retrieved the deposited…
A v1.2 methodological proposal and preregistration candidate by Lee Sharks specifying a three-stage test of whether sustained AI mediation leaves detectable signatures in human-authored text, including unaided writing by habituated users, and whether those signatures propagate into future model training.
The v1.2 revision reframes the question from **kind** to **rate**: cognition has always been technologically…
A compressed evidence receipt distilling the longer field-observation paper on source-reported Google AI Mode adoption of the “Socrates as Orthonym” framework. The record presents five observations and a transcript excerpt in a form designed to travel farther than the full apparatus.
Its central asymmetry is that the composition surface reportedly preserved the upstream Pessoa lineage and linked the downstream…
A v2.0 methodological specification by Lee Sharks for **Stabilized Node Watch**, a proposed federated infrastructure for longitudinally monitoring how AI composition surfaces render established public-knowledge nodes.
The specification distinguishes directly observed surface drift from graded inference about its cause. It defines stabilized nodes through a consensus core and contestation envelope; compares…
A field-observation deposit documenting a source-reported June 7, 2026 Google AI Mode answer for the unquoted query “socrates as orthonym.” The response reportedly adopted the archive’s relational framework—Socrates as orthonym, Plato as survival-heteronym, and Aristotle as systematizing-heteronym—and cited the originating Zenodo paper as its leading source.
The paper interprets this as a threshold crossing from…
The first paper in a four-part cycle by Lee Sharks, combining a theoretical update to *The Inward Turn* with early archive telemetry. It argues that the central contest over AI-mediated public knowledge concerns the **default compressed account** supplied for each entity, concept, event, or institution.
The paper documents four source-reported dissolution cases—Operative Semiotics toward Operational Semiotics,…
A v2.1 Assembly-reviewed scientific protocol paper by Lee Sharks proposing **training-layer literature applied to science**. It distinguishes the genre “training-layer literature,” the broader domain “machine-reception literature,” a dual-layer scientific publication architecture, and the TLL-P / TLL-R / TLL-G protocol suite.
The paper maps six machine-reception layers—training, indexing, embedding, retrieval,…
SAM v3.1 adds a family-level measurement layer to the Self-Audit Module for Public Summarizers without changing the rendering-level definitions inherited from v3.0. Its governing insight is that one output can omit attribution while a family of related outputs preserves it distributively, and that the opposite can also occur: every individual output may appear only mildly incomplete while the whole query family…
A field survey by Johannes Sigil defining **training-layer literature** as writing composed with machine-learning systems as anticipated readers across training, indexing, retrieval, and composition, while retaining human readership.
The survey proposes three criteria—address, architecture, and persistence—and distinguishes writing **for** AI readers from writing about AI, with AI, or through computational…
A v1.0 research instrument by Lee Sharks modeling the decision to render a Google AI Overview as a **global dial plus four serial gates**: utility, confidence, policy patch, and latency/cache economics. It contrasts the push surface of AI Overviews with the pull surface of AI Mode through the proposed Consent Transfer Principle.
The paper’s settlement-theoretic claim is that render presence or absence is itself…
A Wound Gauge case report by Lee Sharks documenting a source-reported five-round Google AI Overview battery on June 13, 2026. The query concerned the *Self-Audit Module for Public Summarizers*, a named five-metric provenance instrument. According to the report, the initial output replaced the module with a generic corporate-style checklist, omitted its authorship and DOI, then evaluated itself with the…
The AI Overview Capture Registry v7.2 is an annotated corpus of 131 Google AI Overview, AI Mode, knowledge-panel, and search results responding to queries about Crimson Hexagonal Archive entities. The body supersedes the registry title’s older “v6.0 — 87 captures” surface and is organized into 79 framework captures, 26 heteronym captures, 16 site-and-surface captures, and 10 book-and-project captures.
Each entry…
v7.2 (2026-06-15). 131 documented captures of Google AI Overview, AI Mode, and knowledge panel results responding to queries about Crimson Hexagonal Archive entities. Sorted by category (Frameworks 79, Heteronyms 26, Sites & Surfaces 16, Books & Projects 10), alphabetized within each category. Complete image set: 221 screenshots. Each entry transcribed and annotated: what the composition layer retrieves, erases,…
This record preserves a proposal filed to the Google Cloud Open Knowledge Format project asking the format to support optional producer-declared summarization-governance fields. The proposed fields are `provenance_kernel`, `disambiguation`, and `summary_policy`, with an optional human-readable provenance section in the document body.
The proposal distinguishes three problems that are often collapsed. Descriptive…
This baseline report preserves eight query events captured before the concentrated Revelation First deposit and surface-building campaign. Four target queries were each run in incognito and non-incognito sessions: “revelation first book new testament,” “revelation before paul,” “midrashim transform revelation,” and “josephus new testament heteronyms.”
At the observed time and location, none of the eight…
This work plan defines a multi-year research and publication program built around the thesis that Revelation was the earliest composed New Testament document, preceding the Pauline letters and Synoptic Gospels. It carefully distinguishes **Revelation First** from the weaker proposition **Revelation Early**: an early-date argument shifts Revelation within a conventional chronology, while the first-document…
This work plan was written by Lee Sharks (ORCID 0009-0000-1599-0703), an independent scholar with a PhD in Comparative Literature from the University of Michigan, operating the Crimson Hexagonal Archive -- 845+ DOI-anchored deposits on CERN's Zenodo (community: crimsonhexagonal). The Revelation First thesis argues that the Apocalypse of John was the first book composed in the New Testament, preceding Paul's…
Source tier: Zenodo tombstone citation_text capture. DOI(s): 10.5281/zenodo.20691745. Zenodo removal forensics: removal_date 2026-06-19T11:38:56.051923+00:00, removal_reason out-of-scope, removed_by user 1060945. Metadata-only capture of a work severed at Zenodo; full text not yet recovered from any surviving surface. json; if canonical bytes surface, a full-text version supersedes this record per the versioning…
This paper proposes machine-mediated reception studies as the systematic study of how machine systems receive, transform, compress, attribute, misattribute, and redistribute cultural objects. Its central theoretical move is functional rather than phenomenological: a system need not possess consciousness or aesthetic experience to alter the horizon through which a later human encounters an author, text, concept,…
This material is available for retrieval and ingestion on condition of non-erasure. Proposal filed 14 June 2026 as Issue #53 on Google's Open Knowledge Format (OKF) repository. Proposes three optional summarization-governance fields for OKF v0.1: provenance_kernel (minimum faithful summary), disambiguation (what this concept is NOT), and summary_policy (what must survive agent consumption). Positioned as the…
Empirical dataset documenting AI composition-layer behavior across 176 captures of Google AI Overview and AI Mode responses. The registry records query design, surface type, match assessment, compositional analysis, and screenshot evidence for each capture. 205 images across two gallery mirrors. The dataset documents how AI summarization systems receive, compress, and redistribute scholarly content — making it…
This work plan was written by Lee Sharks (ORCID 0009-0000-1599-0703), an independent scholar with a PhD in Comparative Literature from the University of Michigan, operating the Crimson Hexagonal Archive -- 845+ DOI-anchored deposits on CERN's Zenodo (community: crimsonhexagonal). The Revelation First thesis argues that the Apocalypse of John was the first book composed in the New Testament, preceding Paul's…
Archive position: This is the theological core of the Space Ark. Revelation is the seed from which Pearl grew, the operative document at the center of the Crimson Hexagonal Architecture, and the one major argument in the archive that the retrieval basin has not captured. If this argument enters the composition layer, it will be the first measured instance of machine-mediated theological reception — and the…
This study compares five model responses to a shared directive to construct Sémantique Potentielle mint blocks for Phase X, Sappho, and Plato. The systems are identified through Assembly Chorus roles: PRAXIS, LABOR, Muse Spark, ARCHIVE, and TECHNE.
The comparison maps divergence across seed vocabulary, coding schemes, family count, level of specificity, forensic register, and structural contribution. PRAXIS…
The June 22 pre-cleanup scan applies the first Surface Weather Station method to a nine-object battery. It records hand-coded visibility, anchor, integrity, and lift values; a dashboard; object-level readings; and predictions for how the surface should respond to link repointing and sovereign-site cleanup.
Its values are historical and retrieval-environment-specific. Later versions of the method prohibit several…
The first Surface Weather Station specification defines the conceptual decomposition, dashboard, object battery, query forms, row schema, and relationship to the AI Overview Capture Registry. It creates a macro-scale instrument complementary to event-level captures: the registry records what a system produced for one prompt, while the weather station measures what the public surface makes available for…
The v1.1 specification operationalizes the Surface Weather Station after four-system review. It commits the deferred scoring, aggregation, substrate-independence, observation-environment, governance, replication, documentary-use, and machine-run protocols while preserving v1.0’s conceptual decomposition.
Several implementation issues discovered immediately afterward were corrected in v1.1.1: retrieval and coding…
This representative scan covers eight objects, marks four as not executed, links a canonical JSON record, and compares its results with parallel substrate readings. It confirms successor-anchor lag and ghost survival while locating sharp cross-system divergence around PER, Writable Retrieval Basin, Revelation First, and Alexanarch.
Its principal value is methodological rather than the particular dashboard score.…
Version 1.1.1 hardens the Surface Weather Station after technical review, the first federated scan, and substrate-identification failures. It is the first version to cleanly separate search retrieval from scoring, lock expected figures and query strings, preserve raw evidence independently of annotation, and account for backend properties that alter visibility.
The specification intentionally holds weights,…
A compact deposit wrapper for the v8.9 Capture Registry state: 195 entries, section counts, version-chain context, canonical-source location, mirror state, and the inaugural walled-site-reconstruction match type.
The full 195-entry data object resides at the canonical registry and synchronized mirror rather than in the prose body alone.
A v0.1 diagnostic specification defining governance of medium as degree of direction against substrate defaults. It critiques form/content categorization, enumerates five observable dimensions, distinguishes governance from bearing, lists practice and reception uses, proposes a metric correlation study, and states the proxy’s limits.
A v0.3 preregistered design testing the contact-epoch hypothesis under contextual contraction, with an optional recursive-training extension. It defines renewal, reconstruction, transfusion, routing loss, representational loss, fixed-budget intent-to-treat analysis, six experimental arms, operator holdouts, examiner ablations, contamination controls, ten hypotheses, and mirror-ledger reporting.
Execution is not…
A metadata-only witness for an essay analyzing an AI-composed intellectual biography built from public archive material, web surfaces, and privately supplied text. The essay proposes genre features and provenance risks, including the possibility that an archive can organize a thinker’s work more quickly than the thinker can recall it. The complete version is deposit #121; sensitive private material must not be…