[ { "id": "O1665-01", "dep": 1665, "claim": "The disclosed field B is the sources the layer surfaced; cards show a source was surfaced, not what was read", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "§0.4", "quote": "The **disclosed field** B is the set of sources the layer surfaced for the address, by card or by name in its body. The cards are **surface indicators**: a card shows that a source was surfaced; it does not show what was retrieved, which passages were read, or what the composer drew from its own parameters.", "kernel": true, "const": false, "lineage": "the disclosed field and its cards", "date": "2026-10-05", "quote_ok": true }, { "id": "O1665-02", "dep": 1665, "claim": "T against L(B) is representational, not a causal test of the layer", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§0.5", "quote": "L(B) reads the full texts of the disclosed sources; what the layer had of them is unknown. The comparison shows what the disclosed field contained that the transcript did not compose, and what the transcript composed that the field does not contain. It is not a causal test of the layer's composition.", "kernel": true, "const": false, "lineage": "T against L(B) is representational", "date": "2026-10-05", "quote_ok": true }, { "id": "O1665-03", "dep": 1665, "claim": "AI Overview is the first surface, chosen as the most exclusive", "kind": "definition", "modality": "stipulation", "strata": [ "D" ], "locus": "§0.5", "quote": "The first stage of the panel is one surface, AI Overview, chosen as the most exclusive.", "kernel": false, "const": false, "lineage": "AI Overview as the most exclusive surface", "date": "2026-10-05", "quote_ok": true }, { "id": "O1665-04", "dep": 1665, "claim": "The first question: what the layer leaves out of the sources it surfaces", "kind": "definition", "modality": "stipulation", "strata": [ "D" ], "locus": "§0.3", "quote": "**First:** what does the composition layer leave out of the sources it itself surfaces (T against L(B))? This question needs no archive and can be asked at any address.", "kernel": false, "const": false, "lineage": "T against L(B) is representational", "date": "2026-10-05", "quote_ok": true }, { "id": "O3-01", "dep": 3, "claim": "A semantic address is the pair (string, surface)", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "_surface_address_rule (registry v12.83)", "quote": "GOING FORWARD, A SEMANTIC ADDRESS IS THE PAIR (STRING, SURFACE).", "kernel": false, "const": false, "lineage": "the Capture Registry", "date": "2026-10-06 (registry v12.83; deposit #3 dated 2026-06-17)", "quote_ok": true }, { "id": "O3-02", "dep": 3, "claim": "Quoting is the decisive measured variable: operative semiotics 5 of 5 archive cards quoted, 1 of 8 unquoted", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "_rerun_alt_rule (registry v12.83)", "quote": "QUOTING IS THE DECISIVE MEASURED VARIABLE IN THIS CORPUS — «operative semiotics» held 5 of 5 archive cards quoted and 1 of 8 unquoted, and by 13 August it resolved ONLY under the exact-phrase operator.", "kernel": true, "const": false, "lineage": "quoting decides admission", "date": "2026-10-06 (registry v12.83; deposit #3 dated 2026-06-17)", "quote_ok": true }, { "id": "O3-03", "dep": 3, "claim": "A regeneration test: 12 of 17 identical, 5 of 17 divergent within three minutes", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "_regeneration_finding (registry v12.83)", "quote": "12 of 17 identical, 5 of 17 DIVERGENT across a ~3 minute gap", "kernel": false, "const": false, "lineage": "the composition regenerates", "date": "2026-10-06 (registry v12.83; deposit #3 dated 2026-06-17)", "quote_ok": true }, { "id": "O3-04", "dep": 3, "claim": "Frame and paste as two views of one observation is falsified for 29% of that sitting", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "_regeneration_finding (registry v12.83)", "quote": "The corpus has always treated frame and paste as two views of ONE observation. That is falsified for 29% of this sitting.", "kernel": false, "const": false, "lineage": "the composition regenerates", "date": "2026-10-06 (registry v12.83; deposit #3 dated 2026-06-17)", "quote_ok": true }, { "id": "O3-05", "dep": 3, "claim": "PER 0.25 on a capture whose finding is liquidation: attribution and binding are independent axes", "kind": "finding", "modality": "documented", "strata": [ "D", "O" ], "locus": "_COMPOSED_AS (registry v12.83)", "quote": "«semantic economy» on 2026-08-13 scored PER 0.25 — three quarters of provenance intact — on a capture whose own finding reads LIQUIDATION, from critical framework to enterprise vendor copy. A registry that records only PER reports that event as healthy.", "kernel": false, "const": false, "lineage": "liquidation beside attribution", "date": "2026-10-06 (registry v12.83; deposit #3 dated 2026-06-17)", "quote_ok": true }, { "id": "O3-06", "dep": 3, "claim": "An entry is a capture only if it shows engagement with sources through a retrieval or composition surface", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "_registers_rule (registry v12.83)", "quote": "an entry is a CAPTURE only if it shows an engagement with archival sources through a retrieval or composition surface", "kernel": false, "const": false, "lineage": "the Capture Registry", "date": "2026-10-06 (registry v12.83; deposit #3 dated 2026-06-17)", "quote_ok": true }, { "id": "O1461-01", "dep": 1461, "claim": "v10.9: 328 addresses, 470 captures, 41 dates, 13 surfaces; 290 scored for PER, mean 0.603", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Description", "quote": "328 semantic addresses carrying 470 captures across 41 distinct dates and thirteen surfaces, 451 of them with a verbatim transcript and 287 with a frame; 290 scored for Provenance Erasure Rate at a mean of 0.603.", "kernel": true, "const": false, "lineage": "the Capture Registry", "date": "2026-08-14", "quote_ok": true }, { "id": "O1461-02", "dep": 1461, "claim": "Surfaces at v10.9: Google AI Overview 343", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "State at v10.9", "quote": "**Surfaces** — Google AI Overview 343 · UNDETERMINED 63 · UNRESOLVED 50 · ChatGPT 3 · Google Knowledge panel 2 · Google Search results page (no AI surface) 2 · Google AI Mode (native) 1 · Google Scholar 1", "kernel": false, "const": false, "lineage": "the Capture Registry", "date": "2026-08-14", "quote_ok": true }, { "id": "O1461-03", "dep": 1461, "claim": "One surface + one semantic address + one date = one capture", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "The citable unit", "quote": "**ONE SURFACE + ONE SEMANTIC ADDRESS + ONE DATE = ONE CAPTURE.**", "kernel": false, "const": false, "lineage": "the Capture Registry", "date": "2026-08-14", "quote_ok": true }, { "id": "O1461-04", "dep": 1461, "claim": "A capture is falsified by a re-run returning materially different composition", "kind": "falsifier", "modality": "self-description", "strata": [ "D" ], "locus": "Falsification Conditions", "quote": "Any capture is falsified by a re-run at the same address, surface, authentication state and date returning materially different composition, which is why the exact issued string and the rerun URL are recorded.", "kernel": true, "const": false, "lineage": "the Capture Registry", "date": "2026-08-14", "quote_ok": true }, { "id": "O1461-05", "dep": 1461, "claim": "A measured reclamation: semantic exhaustion composed with the archive's definition on 2026-08-14 against absorption on 2026-07-06", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "What this instrument has measured", "quote": "A measured reclamation, 2026-08-14: «\"semantic exhaustion\"» returning the archive's definition and citing EA-MPAI-SEMEX-01 by name, against 2026-07-06 when the coinage was absorbed by semantic satiation across seven of seven third-party sources.", "kernel": false, "const": false, "lineage": "supply-side composition", "date": "2026-08-14", "quote_ok": true }, { "id": "O1643-01", "dep": 1643, "claim": "As of 27 September 2026 the registry holds 663 answers at 493 addresses", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Part Two", "quote": "As of 27 September 2026 the registry holds **663 recorded answers at 493 search addresses**, from Google AI Overview, Google AI Mode, ChatGPT, Grok, Bing Copilot, Perplexity, Claude and Qwen.", "kernel": false, "const": false, "lineage": "the Capture Registry", "date": "2026-09-27", "quote_ok": true }, { "id": "O1614-01", "dep": 1614, "claim": "Transcript coverage: 923 of 957 citable units carry machine text, 96.45%", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§2", "quote": "923 have their machine text. 34 do not. Coverage 96.45%.", "kernel": false, "const": false, "lineage": "the Capture Registry", "date": "2026-09-15", "quote_ok": true }, { "id": "O833-01", "dep": 833, "claim": "The AI summarizer functions as a horizon of reception", "kind": "claim", "modality": "interpretation", "strata": [ "D", "O" ], "locus": "Abstract", "quote": "Its central claim is that the AI summarizer now functions as a *horizon of reception*: for many users, the machine-composed summary becomes the first effective encounter with a cultural object, author, concept, or tradition.", "kernel": true, "const": true, "lineage": "the summarizer as horizon of reception", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-02", "dep": 833, "claim": "In 131 captures, six recurrent patterns appear (empirical claim, narrow)", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§1 Claims Hierarchy", "quote": "The *empirical claim* is narrower: in 131 documented captures of Google AI Overview, AI Mode, and knowledge-panel responses to Crimson Hexagonal Archive entities, six recurrent patterns appear.", "kernel": false, "const": false, "lineage": "six patterns of machine-mediated reception", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-03", "dep": 833, "claim": "Composition layer: the pipeline that synthesizes from retrieved sources; the index finds, the composition reads", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "§3", "quote": "**Composition layer.** The computational pipeline that synthesizes a summary from retrieved sources. Distinguished from the search index, which retrieves and ranks documents. The composition layer reads; the search index finds.", "kernel": false, "const": false, "lineage": "composition layer defined", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-04", "dep": 833, "claim": "Compositional bystanding: the archive's record at organic rank #1, not cited by the Overview", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§5.3", "quote": "In the semantic-commodity-form capture, the archive's Ratification Record appears at organic search rank #1 but is not cited as a composition source by the AI Overview.", "kernel": false, "const": false, "lineage": "ranked first, not cited", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-05", "dep": 833, "claim": "Inferred: organic index and Overview pipeline are separate systems with different admission criteria; cause inferred", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§5.3", "quote": "This pattern reveals a crucial architectural feature: the organic search index and the AI Overview composition pipeline are separate systems with different admission criteria.", "kernel": false, "const": false, "lineage": "ranked first, not cited", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-06", "dep": 833, "claim": "The pattern is documented; its cause is inferred", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§5.3", "quote": "Whether this exclusion reflects a deliberate content policy, a relevance-model threshold, or an artifact of the pipeline's passage-selection heuristics is not determinable from output observation alone. The pattern is documented; its cause is inferred.", "kernel": false, "const": false, "lineage": "ranked first, not cited", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-07", "dep": 833, "claim": "Entity substitution: Lee Sharks resolved toward Mary Lee the shark", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§5.2", "quote": "**Entity substitution.** \"Lee Sharks\" is resolved toward \"Mary Lee,\" a tagged great white shark tracked by OCEARCH. The composition layer prefers the more populous real-world entity.", "kernel": false, "const": false, "lineage": "entity substitution at the author's name", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-08", "dep": 833, "claim": "Temporal drag: the layer serves pre-correction versions after the sources are corrected", "kind": "finding", "modality": "interpretation", "strata": [ "D" ], "locus": "§5.5", "quote": "The composition layer locks onto pre-correction cached representations and serves outdated versions long after the source repositories have versioned past them.", "kernel": false, "const": false, "lineage": "temporal drag", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-09", "dep": 833, "claim": "Epistemic hedging: the archive's reading described as an esoteric, fringe theory", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§5.6", "quote": "In the Josephus Thesis capture (#92), the composition layer describes the archive's reading as an \"esoteric, fringe theory.\"", "kernel": false, "const": false, "lineage": "epistemic hedging", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-10", "dep": 833, "claim": "A minority reading of Sappho 31 developed in the archive presented as the default in AIO", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Abstract", "quote": "In one case, a minority scholarly interpretation of Sappho Fragment 31 — developed within the archive — is now presented as the default understanding in Google's AI Overview.", "kernel": false, "const": false, "lineage": "the archive's reading composed as default", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-11", "dep": 833, "claim": "Self-capture: the layer accurately described the Capture Registry (June 2026)", "kind": "finding", "modality": "documented", "strata": [ "D", "R" ], "locus": "§7", "quote": "Entry #1 in the sorted registry (\"AI overview capture registry\") documents the capture registry capturing itself: the composition layer accurately describes the AI Overview Capture Registry, its methodology, and its purpose.", "kernel": false, "const": false, "lineage": "the registry at its own address", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-12", "dep": 833, "claim": "Limit: a single-archive corpus", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§8", "quote": "**Single-archive corpus.** The dataset studies machine reception of the Crimson Hexagonal Archive, not all cultural objects.", "kernel": true, "const": false, "lineage": "the summarizer as horizon of reception", "date": "2026-06-15", "quote_ok": true }, { "id": "O833-13", "dep": 833, "claim": "Non-claims: not all systems; the captures do not reveal Google's internal mechanisms", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§9", "quote": "**This paper does not claim:** - That this dataset represents all AI systems or all composition layers. - That the captures reveal Google's internal mechanisms (the patterns are inferred from observable outputs).", "kernel": false, "const": false, "lineage": "the summarizer as horizon of reception", "date": "2026-06-15", "quote_ok": true }, { "id": "O1459-01", "dep": 1459, "claim": "Compositional entity liquidation: the address survives, the composed referent disappears", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "Abstract", "quote": "This packet classifies that reception state as **compositional entity liquidation**, a proposed severe subtype of provenance erasure: the evidentiary address survives, but the composed referent disappears while its semantic material remains active elsewhere.", "kernel": true, "const": false, "lineage": "compositional entity liquidation", "date": "2026-08-13", "quote_ok": true }, { "id": "O1459-02", "dep": 1459, "claim": "On 13 August 2026 the registry's own name no longer appeared in broad-match composition", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Abstract", "quote": "On 13 August 2026, the registry became the object of the phenomenon it had already been built to measure. In the observed broad-match composition condition, the named entity **AI Overview Capture Registry** no longer appeared in composition.", "kernel": false, "const": false, "lineage": "the registry at its own address", "date": "2026-08-13", "quote_ok": true }, { "id": "O1459-03", "dep": 1459, "claim": "Google AI Overview is the observed surface; the registry is the instrument; never collapsed", "kind": "relation", "modality": "stipulation", "strata": [ "R" ], "locus": "§10.6", "quote": "**Google AI Overview** is an observed product surface. **AI Overview Capture Registry** is the research instrument observing it. The two must never be collapsed.", "kernel": true, "const": false, "lineage": "instrument and surface kept apart", "date": "2026-08-13", "quote_ok": true }, { "id": "O1459-04", "dep": 1459, "claim": "Generic explainers of AIO may describe the platform, not the instrument observing it", "kind": "relation", "modality": "interpretation", "strata": [ "R" ], "locus": "§11.4", "quote": "The observed source set also contains generic explanatory material about what Google AI Overviews are and how they affect search behavior. Such pages may accurately describe the **observed platform**. They do not describe the **instrument observing it**.", "kernel": false, "const": false, "lineage": "instrument and surface kept apart", "date": "2026-08-13", "quote_ok": true }, { "id": "O1612-01", "dep": 1612, "claim": "Presented sets on AIO: tied addresses admit the archive at 0.74–0.86, untied at 0.26–0.53", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§3", "quote": "The presented set on Google AI Overview, by arm 195 observations, June through September. Tied addresses — the entity or a heteronym in the issued string — admit the archive at a mean share of 0.74 to 0.86 in every month with two zero-archive observations out of 101. Untied addresses admit it at 0.26 to 0.53 with sixteen zero-archive observations out of 94.", "kernel": true, "const": false, "lineage": "tied and untied admission", "date": "2026-09-14", "quote_ok": true }, { "id": "O1612-02", "dep": 1612, "claim": "The profile cannot separate absence from the organic field, rejection at card admission, and use in composition", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§2", "quote": "It cannot separate three phenomena that the composition-layer measures collapse: the archive absent from the organic field; the archive strong organically and rejected at card admission; the archive admitted to the cards and ignored or recoded in composition.", "kernel": true, "const": false, "lineage": "tied and untied admission", "date": "2026-09-14", "quote_ok": true }, { "id": "O1612-03", "dep": 1612, "claim": "Tied arm narrows: fewer sources, longer answers, fewer external sources", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§3", "quote": "The tied arm narrows: sources per observation fall from five in July–August to two or three in the 11 September re-measurements while the answers lengthen — 169 to 650 words at *alexanarch symbolon*, 57 to 386 at *rebekah cranes sappho* — and external sources per tied observation fall from 1.7 to 0.6.", "kernel": false, "const": false, "lineage": "tied and untied admission", "date": "2026-09-14", "quote_ok": true }, { "id": "O1612-04", "dep": 1612, "claim": "Untied substitutes become institutional: scholarly platforms, Wikipedia, YouTube, code hosts", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§3", "quote": "The untied arm's substitutes change character: the external cards admitted at concept addresses are \"other web\" in June and, by August, scholarly platforms (57), Wikipedia (28, from 1), YouTube (16) and code hosts (15).", "kernel": false, "const": false, "lineage": "tied and untied admission", "date": "2026-09-14", "quote_ok": true }, { "id": "O1612-05", "dep": 1612, "claim": "Four ranked-first-and-uncited observations are exemplars, not a rate", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§2", "quote": "The four observations in which a capture's prose records the archive ranked first organic and uncited are exemplars of the second, not a rate.", "kernel": false, "const": false, "lineage": "ranked first, not cited", "date": "2026-09-14", "quote_ok": true }, { "id": "O1612-06", "dep": 1612, "claim": "The registry records the presented card set on 314 of 588 observations", "kind": "finding", "modality": "documented", "strata": [ "D", "O" ], "locus": "§1", "quote": "A composition surface that renders an Overview presents its sources — inline chips and a carousel. The Capture Registry records that set as `cite_list` on 314 of its 588 observations.", "kernel": false, "const": false, "lineage": "the Capture Registry", "date": "2026-09-14", "quote_ok": true }, { "id": "O187-01", "dep": 187, "claim": "Whether an AIO appears is the output of a stacked decision system", "kind": "claim", "modality": "hypothesis", "strata": [ "D" ], "locus": "§0", "quote": "Whether an AI Overview (AIO) appears for a given query is not noise and not mystery: it is the output of a stacked decision system whose components are separately documented — in Google's own patents, in its public statements, in its emergency interventions, and in a now-substantial third-party measurement literature.", "kernel": true, "const": true, "lineage": "the render decision as gate stack", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-02", "dep": 187, "claim": "Consent Transfer Principle: the threshold varies inversely with the explicitness of the request for synthesis", "kind": "definition", "modality": "hypothesis", "strata": [ "D", "O" ], "locus": "§I", "quote": "Consent Transfer Principle: *for a generative search surface, the rendering threshold varies inversely with the explicitness of the user's request for synthesis.*", "kernel": false, "const": false, "lineage": "the render decision as gate stack", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-03", "dep": 187, "claim": "Otterly: 100 queries triggered AI Mode 100 times, an AI Overview 49 times", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§I", "quote": "In the cleanest controlled comparison to date (Otterly, April 2026), the same 100 queries triggered AI Mode 100 times and an AI Overview 49 times.", "kernel": false, "const": false, "lineage": "when it appears", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-04", "dep": 187, "claim": "Where both fire, AI Mode ~310 citations against the AIO's ~51; AIOs over-index official sources", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§I", "quote": "where both fire on the same query, AI Mode averages roughly 310 citations against the AIO's ~51, drawing on ~3,600 unique domains against ~615, via the query fan-out technique (parallel sub-queries synthesized into one response). And the pools are differently shaped: AIOs over-index official and governmental sources — institutional consensus", "kernel": false, "const": false, "lineage": "which sources it cites", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-05", "dep": 187, "claim": "Patent: a method of selecting none, one or multiple generative models; none is a claimed outcome", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§III.1", "quote": "in Google's patent family \"Generative summaries for search results\" (US11769017B1, filed March 2023, with continuations US11886828B1, US11900068B1, US12118325). Three features of the family bear directly on triggering: 1. The selection step exists by name. The specification describes a method of \"selecting none, one, or multiple generative model(s) to utilize in generating response(s) to render responsive to a query.\" *None* is a claimed outcome.", "kernel": false, "const": false, "lineage": "the render decision as gate stack", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-06", "dep": 187, "claim": "Patent: the summary can include content not derivable from the processed content", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§III.3", "quote": "The specification states the generated summary \"can also include content that is not directly (or even indirectly) derivable from the content processed using the LLM,\" drawn from the model's world knowledge.", "kernel": false, "const": false, "lineage": "content beyond the sources", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-07", "dep": 187, "claim": "Patent continuations: different submissions of the same query yield different summaries", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§III.2", "quote": "The continuations specify that differing \"additional content\" is processed for different submissions of the same query — by query location, query language, and attributes of the user profile — so that *different submissions of the same query yield different summaries*, deliberately, to make the summary \"resonate with the user that submitted the query.\"", "kernel": false, "const": false, "lineage": "the composition regenerates", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-08", "dep": 187, "claim": "January 2026 health removals: near-variant phrasings continued to trigger summaries", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§II", "quote": "In January 2026, following a Guardian investigation into harmful health renderings, Google removed AIOs for specific medical queries — while near-variant phrasings of the same queries continued to trigger summaries.", "kernel": false, "const": false, "lineage": "health removals, January 2026", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-09", "dep": 187, "claim": "Read as decisive: the emergency layer operates on query strings, not meanings", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§II", "quote": "That detail is forensically decisive: the emergency layer operates on query strings, not meanings.", "kernel": false, "const": false, "lineage": "health removals, January 2026", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-10", "dep": 187, "claim": "Composition churns: cited URLs change about twelve times a month per keyword", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§II", "quote": "cited URLs change on the order of twelve times per month per keyword; 96% of stable-trigger keywords saw domain swaps within a month; 91% of URLs fall out, fewer than half return.", "kernel": false, "const": false, "lineage": "the composition regenerates", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-11", "dep": 187, "claim": "Google: systems determine where an AI response is useful, not always 100 percent consistent", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§II", "quote": "Google has stated that for some topics — current events named explicitly — both surfaces may return a list of links rather than a generated response, and that the systems \"automatically determine where an AI response will be useful,\" conceding it is \"not always 100 percent consistent.\"", "kernel": false, "const": false, "lineage": "when it appears", "date": "2026-06-11", "quote_ok": true }, { "id": "O187-12", "dep": 187, "claim": "Prediction P4: post-incident suppressions fail on near-variant phrasings", "kind": "prediction", "modality": "hypothesis", "strata": [ "D" ], "locus": "§V", "quote": "P4 (string-local patches): suppressions installed after public incidents fail on near-variant phrasings at rates inconsistent with semantic gating — testable against any future emergency removal.", "kernel": false, "const": false, "lineage": "health removals, January 2026", "date": "2026-06-11", "quote_ok": true }, { "id": "O597-01", "dep": 597, "claim": "Invisibly invisible governance: semantic modulation variably distributed, hard to localize, deniable", "kind": "definition", "modality": "stipulation", "strata": [ "D", "O" ], "locus": "Abstract", "quote": "Through an immanent phenomenological analysis of real search interfaces—especially Google Search, AI Overviews, and related retrieval surfaces—we develop the concept of **invisibly invisible governance**: a mode of platform power in which acts of semantic modulation occur in ways that are variably distributed, difficult to localize, and publicly deniable.", "kernel": true, "const": false, "lineage": "invisibly invisible governance", "date": "2026-03-23", "quote_ok": true }, { "id": "O597-02", "dep": 597, "claim": "Google states AIO appears when its systems determine it will be especially helpful", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§3", "quote": "Google publicly states that AI Overviews appear when its systems determine that generative AI will be especially helpful, and that they often do not trigger.", "kernel": false, "const": false, "lineage": "when it appears", "date": "2026-03-23", "quote_ok": true }, { "id": "O597-03", "dep": 597, "claim": "No public log of why the threshold was crossed", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§3", "quote": "When the Overview does not appear, this absence is not explained as an event. It is experienced as ordinary search. The interface gives no public log of why the threshold was crossed yesterday and not today.", "kernel": false, "const": false, "lineage": "invisibly invisible governance", "date": "2026-03-23", "quote_ok": true }, { "id": "O597-04", "dep": 597, "claim": "Health restriction implemented as a silent withdrawal of the format", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§3", "quote": "In multiple documented cases, Overviews generated unreliable or dangerous health information, prompting Google to restrict Overview triggering for certain query classes. This restriction was implemented not as a visible policy announcement attached to each query but as a silent withdrawal of the Overview format.", "kernel": false, "const": false, "lineage": "health removals, January 2026", "date": "2026-03-23", "quote_ok": true }, { "id": "O597-05", "dep": 597, "claim": "The paper cites Google's Help page answer 14901683", "kind": "relation", "modality": "documented", "strata": [ "R" ], "locus": "note 15", "quote": "[^15]: Google, \"About AI Overviews,\" https://support.google.com/websearch/answer/14901683 (accessed 2026).", "kernel": false, "const": false, "lineage": "the same Help page", "date": "2026-03-23", "quote_ok": true }, { "id": "O64-01", "dep": 64, "claim": "Under the AI Overview the engine composes the answer itself", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§III", "quote": "Under the link settlement, the engine composed an ordering and the author composed what was ordered. Under the AI Overview, the engine composes the answer itself. The user reads the settlement's composition, not the source authors'.", "kernel": true, "const": false, "lineage": "the compositional settlement", "date": "2026-04-18", "quote_ok": true }, { "id": "O64-02", "dep": 64, "claim": "The UI of the AI Overview is the definite article made architectural", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§III", "quote": "The UI of the AI Overview is the definite article made architectural. It physically instantiates *the*: this is the answer, rendering the plurality of the web implicit.", "kernel": false, "const": false, "lineage": "the compositional settlement", "date": "2026-04-18", "quote_ok": true }, { "id": "O64-03", "dep": 64, "claim": "AI Overviews announced at I/O on May 14, 2024", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§III", "quote": "Google announced AI Overviews at I/O on May 14, 2024, and rolled them out progressively thereafter.", "kernel": false, "const": false, "lineage": "launch, May 2024", "date": "2026-04-18", "quote_ok": true }, { "id": "O115-01", "dep": 115, "claim": "Identity in Search is admission: an entity can be indexed and retrievable yet not compositionally real", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "Abstract", "quote": "The central claim is that identity in Google Search is not merely recognition but admission: an entity may be known, indexed, and organically retrievable while still failing to become compositionally real in AI Overview or AI Mode.", "kernel": true, "const": false, "lineage": "admission to composition", "date": "2026-05-21", "quote_ok": true }, { "id": "O115-02", "dep": 115, "claim": "Read as an external specification, not as how Google works internally", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "Confidence Labels", "quote": "The document should be read as an external specification: not “this is exactly how Google works internally,” but “this is the most complete outside-view architecture presently defensible from public documentation, leak-derived signals, and empirical behavior.”", "kernel": true, "const": false, "lineage": "admission to composition", "date": "2026-05-21", "quote_ok": true }, { "id": "O742-01", "dep": 742, "claim": "Hypothesis: a single-owner discount", "kind": "claim", "modality": "hypothesis", "strata": [ "D", "O" ], "locus": "Abstract", "quote": "a *single-owner discount* under which an internally dense body of work attributed to one provenance owner is treated as weakly corroborated, regardless of internal coherence or factual quality, while equivalent content distributed across multiple institutional owners is treated as authoritative.", "kernel": true, "const": false, "lineage": "the single-owner discount", "date": "2026-05-19", "quote_ok": true }, { "id": "O742-02", "dep": 742, "claim": "A document may be indexed and snippet-eligible yet excluded from composition", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "Abstract", "quote": "Generative search systems — Google's AI Overviews, AI Mode, and equivalent systems — operate through a stratified pipeline in which document retrieval and answer composition are governed by separable mechanisms. A document may be indexed, retrievable, and snippet-eligible while remaining systematically excluded from generative composition.", "kernel": false, "const": false, "lineage": "admission to composition", "date": "2026-05-19", "quote_ok": true }, { "id": "O742-03", "dep": 742, "claim": "Architectural and testable, not a disclosure of Google's internal system", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "Abstract", "quote": "The claim is architectural and testable, not a disclosure of Google's internal production system.", "kernel": true, "const": false, "lineage": "the single-owner discount", "date": "2026-05-19", "quote_ok": true }, { "id": "O740-01", "dep": 740, "claim": "Entity-Level Compositional Suppression defined", "kind": "definition", "modality": "stipulation", "strata": [ "D", "O" ], "locus": "Abstract", "quote": "A generative-search system exhibits *Entity-Level Compositional Suppression* (ECS) when it excludes the dominant organic-resolution entity from the AI Overview's admitted source set and composes a substitute answer from less query-responsive entities, despite the correct entity occupying the top organic search results.", "kernel": true, "const": false, "lineage": "entity-level compositional suppression", "date": "2026-05-19", "quote_ok": true }, { "id": "O740-02", "dep": 740, "claim": "Documented by three captures from Google AI Mode on May 19, 2026", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "Abstract", "quote": "This worked example documents three captures from Google AI Mode on May 19, 2026 demonstrating the mechanism.", "kernel": false, "const": false, "lineage": "entity-level compositional suppression", "date": "2026-05-19", "quote_ok": true }, { "id": "O740-03", "dep": 740, "claim": "Compositional substitution: title-exact query, work at organic 1–3, Overview cites three distant substitutes", "kind": "definition", "modality": "stipulation", "strata": [ "D" ], "locus": "§2.3", "quote": "When the user queries the literal title of a deposited primary work attributed to the suppressed entity, the standard search results surface the work at positions 1–3, including the work's own domain, the author's publication of the work, and the DOI-anchored deposit. The AI Overview, drawing from the same retrieval set, constructs an answer from three semantically distant non-primary sources, none of which is the title-matched work.", "kernel": false, "const": false, "lineage": "entity-level compositional suppression", "date": "2026-05-19", "quote_ok": true }, { "id": "O740-04", "dep": 740, "claim": "Publishers' July 2025 complaint: misuse of content, no opt-out short of disappearing from Search", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§7", "quote": "The complaint alleged that Google Search is misusing web content for Google's AI Overviews, causing significant harm to publishers in the form of traffic, readership and revenue loss, and that publishers do not have the option to opt out unless they are willing to disappear from Google search results entirely.", "kernel": false, "const": false, "lineage": "regulators and publisher complaints", "date": "2026-05-19", "quote_ok": true }, { "id": "O740-05", "dep": 740, "claim": "9 December 2025: EC formal antitrust proceedings against Google's AI Overviews", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§7", "quote": "**December 9, 2025.** The European Commission opened formal antitrust proceedings against Google's AI Overviews and YouTube, following the July 2025 complaint, investigating whether Google may have imposed unfair terms on publishers and content creators while placing rival AI model developers at a disadvantage.", "kernel": false, "const": false, "lineage": "regulators and publisher complaints", "date": "2026-05-19", "quote_ok": true }, { "id": "O740-06", "dep": 740, "claim": "February 2026 EPC complaint: AIO in over 40% of informational results; traffic declines over 30%", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§7", "quote": "The complaint cites that AI Overviews appear in more than 40% of search results for informational queries, with independent studies estimating traffic declines of over 30% for affected queries and some publishers reporting click-through reductions exceeding 50% on desktop and mobile.", "kernel": false, "const": false, "lineage": "traffic and clicks", "date": "2026-05-19", "quote_ok": true }, { "id": "O740-07", "dep": 740, "claim": "Entity integrity at the composition layer: a harm category the regulatory record lacks", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§7", "quote": "*Entity integrity at the composition layer.* The AI Overview can exclude an organically dominant entity from compositional reality and substitute a false entity frame, even when the Overview is otherwise functioning as designed (the system did synthesize; it was not in low-confidence mode declining to compose).", "kernel": false, "const": false, "lineage": "entity-level compositional suppression", "date": "2026-05-19", "quote_ok": true }, { "id": "O740-08", "dep": 740, "claim": "No claim of intent", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§10", "quote": "**No claim of intent.** This document does not claim that Google has made conscious decisions to suppress the Lee Sharks entity.", "kernel": true, "const": false, "lineage": "entity-level compositional suppression", "date": "2026-05-19", "quote_ok": true }, { "id": "O740-09", "dep": 740, "claim": "The captures are diagnostic of one observed instance, not of universal application", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§10", "quote": "The captures presented are diagnostic of the mechanism's operation in at least one observed instance, not of its universal application.", "kernel": false, "const": false, "lineage": "entity-level compositional suppression", "date": "2026-05-19", "quote_ok": true }, { "id": "O167-01", "dep": 167, "claim": "A parallel AIO response on desktop rendered the framework structurally identically to AI Mode", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§1.2", "quote": "A parallel Google AI Overview response to the same query, captured separately from a desktop search, produced a structurally identical rendering with two differences of inflection", "kernel": true, "const": false, "lineage": "framework adopted, author excised", "date": "2026-06-08", "quote_ok": true }, { "id": "O167-02", "dep": 167, "claim": "Suggests both surfaces draw on the same compositional grounding", "kind": "claim", "modality": "hypothesis", "strata": [ "D" ], "locus": "§1.2", "quote": "The two surfaces — Google AI Mode (mobile) and Google AI Overview (desktop) — render the framework with consistent structural commitments and minor surface variation, which suggests both are drawing on the same underlying compositional grounding rather than producing independent paraphrases.", "kernel": false, "const": false, "lineage": "framework adopted, author excised", "date": "2026-06-08", "quote_ok": true }, { "id": "O164-01", "dep": 164, "claim": "The composition layer attributes the upstream lineage, links the downstream artifact, elides the present author", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "The Asymmetry as Structural Operation", "quote": "The composition layer does not refuse attribution generically. It attributes the *concept lineage* (upstream) and surfaces the *artifact* (downstream Zenodo deposit). What it elides is the *present author*", "kernel": true, "const": false, "lineage": "framework adopted, author excised", "date": "2026-06-08", "quote_ok": true }, { "id": "O716-01", "dep": 716, "claim": "Provenance Erasure Rate: share of source-dependent claims presented without attribution", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "Abstract", "quote": "This paper introduces **Provenance Erasure Rate (PER)** as a complementary metric: the proportion of source-dependent claims in an AI-composed output that are presented without explicit attribution.", "kernel": true, "const": false, "lineage": "provenance erasure rate", "date": "2026-05-03", "quote_ok": true }, { "id": "O716-02", "dep": 716, "claim": "An AIO constructed a false biography of a living author from fragments of his poetry", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Abstract", "quote": "A motivating case study documents a Google AI Overview that constructed a false biography of a living author from real fragments in the author's published poetry: the fragments survived compression, but their provenance and meaning did not.", "kernel": false, "const": false, "lineage": "provenance erasure rate", "date": "2026-05-03", "quote_ok": true }, { "id": "O198-01", "dep": 198, "claim": "The Self-Audit Module dissolved in AIO: attribution stripped, metrics fabricated, self-scored", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Abstract", "quote": "The finding is total dissolution: the Overview retrieved the module, absorbed its semantic content, stripped every attribution marker, repackaged the specification as generic industry advice, fabricated replacement metrics, ran the fabricated metrics on itself, gave itself perfect scores, and required five rounds of author intervention to surface the actual instrument.", "kernel": true, "const": false, "lineage": "the provenance-erasure specification erased", "date": "2026-06-13", "quote_ok": true }, { "id": "O1643-02", "dep": 1643, "claim": "Invented details and quotations: facts or quoted words in none of the sources", "kind": "definition", "modality": "stipulation", "strata": [ "D", "O" ], "locus": "Group B §6", "quote": "**What it is.** The answer adds specific facts, or puts words in quotation marks, that appear in none of its sources.", "kernel": true, "const": false, "lineage": "content beyond the sources", "date": "2026-09-27", "quote_ok": true }, { "id": "O1643-03", "dep": 1643, "claim": "phase x marx: the 1936 Soviet formula quoted as Marx's", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Group B §6", "quote": "**\"phase x marx\"**, 27 September 2026: \"From each according to his ability, to each according to his work,\" attributed to Marx. It is the 1936 Soviet constitutional formula.", "kernel": false, "const": false, "lineage": "content beyond the sources", "date": "2026-09-27", "quote_ok": true }, { "id": "O1643-04", "dep": 1643, "claim": "heteronyms provenance theory: an abbreviation the archive never uses introduced as established", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Group B §6", "quote": "**\"heteronyms provenance theory\"**, Google AI Overview, 14 June 2026: \"Heteronymic Provenance Theory (HPT)\", introduced as an established field with an established abbreviation. The archive uses the phrase seven times and the abbreviation zero times.", "kernel": false, "const": false, "lineage": "content beyond the sources", "date": "2026-09-27", "quote_ok": true }, { "id": "O1643-05", "dep": 1643, "claim": "The essay makes no claim about minds; it records answers, sources and the direction of difference", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "Part One", "quote": "This essay makes no claim about minds. It records what the answers said, what the sources said, and which way the difference runs.", "kernel": true, "const": false, "lineage": "content beyond the sources", "date": "2026-09-27", "quote_ok": true }, { "id": "O1615-01", "dep": 1615, "claim": "Hypothesis: a layer can be porous to a source's ontology and closed to its provenance", "kind": "claim", "modality": "hypothesis", "strata": [ "D", "O" ], "locus": "§0", "quote": "A composition layer can be porous to a source's *ontology* while remaining closed to its *provenance*, and the porosity is graded by representational level.", "kernel": true, "const": false, "lineage": "ontology admitted, provenance closed", "date": "2026-09-15", "quote_ok": true }, { "id": "O1615-02", "dep": 1615, "claim": "The loss recurs at untied addresses, not general: tied addresses transport provenance", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§0", "quote": "The loss is recurrent at untied addresses and not general: at entity-tied addresses in the same record provenance transports well (§1), and on another surface the same week it transported with the instrument (§2a).", "kernel": false, "const": false, "lineage": "ontology admitted, provenance closed", "date": "2026-09-15", "quote_ok": true }, { "id": "O172-01", "dep": 172, "claim": "Where an AIO is present, zero-click reaches 80–83%", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§I", "quote": "Where an AI Overview is present, the zero-click rate reaches 80–83% — four of five askers never leave the answer surface.", "kernel": false, "const": false, "lineage": "traffic and clicks", "date": "2026-06-09", "quote_ok": true }, { "id": "O172-02", "dep": 172, "claim": "Operative semiotics reorganized by AIO around Pearson's Operational Semiotics", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Case 1", "quote": "The archive's discipline-naming term, present in over 120 deposits, was reorganized by Google's AI Overview around Charls Pearson's Theory of Operational Semiotics — a 1990s formal linguistics of sentence-factoring operators with no conceptual overlap and no shared citations.", "kernel": false, "const": false, "lineage": "substitution by a neighbour", "date": "2026-06-09", "quote_ok": true }, { "id": "O172-03", "dep": 172, "claim": "Index cases do not establish prevalence", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§II", "quote": "These four index cases support the asymmetric-dissolution heuristic; they do not establish prevalence.", "kernel": false, "const": false, "lineage": "substitution by a neighbour", "date": "2026-06-09", "quote_ok": true }, { "id": "O1091-01", "dep": 1091, "claim": "Seer: where AIO appears, top organic CTR drops about 79%", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§I", "quote": "Where AI Overviews appear, the click-through rate for the top organic link drops by roughly 79% (Seer Interactive, September 2025, 3,119 queries across 42 organizations, aggregate organic-CTR reduction 61%).", "kernel": true, "const": false, "lineage": "traffic and clicks", "date": "2026-04-15", "quote_ok": true }, { "id": "O1091-02", "dep": 1091, "claim": "The AI Overview does not sit above the results; it is the answer", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§I", "quote": "The AI Overview does not sit above the search results; it *is* the answer.", "kernel": false, "const": false, "lineage": "the compositional settlement", "date": "2026-04-15", "quote_ok": true }, { "id": "O1091-03", "dep": 1091, "claim": "43% of AIO citations self-referential to Google properties", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§II", "quote": "43% of AI Overview citations are self-referential — Google citing Google properties (YouTube, Maps, Shopping, Knowledge Graph panels).", "kernel": false, "const": false, "lineage": "which sources it cites", "date": "2026-04-15", "quote_ok": true }, { "id": "O1091-04", "dep": 1091, "claim": "Ahrefs: only 38% of citations from top-10 pages early 2026, 76% mid-2025", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§II", "quote": "Only 38% of AI Overview citations come from top-10 ranked pages (Ahrefs, early 2026). This was 76% in mid-2025.", "kernel": false, "const": false, "lineage": "citations and organic rank", "date": "2026-04-15", "quote_ok": true }, { "id": "O1091-05", "dep": 1091, "claim": "Wikipedia pageviews down 8% despite over 1.1 million AIO citations", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "§III", "quote": "Wikipedia experienced an 8% year-over-year decline in human pageviews (2025 vs 2024) *despite being cited over 1.1 million times* in AI Overviews.", "kernel": false, "const": false, "lineage": "traffic and clicks", "date": "2026-04-15", "quote_ok": true }, { "id": "O1091-06", "dep": 1091, "claim": "UK CMA proposed a publisher opt-out; Google agreed on 2026-03-19 to explore opt-out controls", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§III", "quote": "the UK Competition and Markets Authority opened a strategic-market-status review and, in January 2026, proposed requiring Google to provide publishers with a meaningful opt-out from AI Overviews without loss of visibility in classical search — Google agreed on 2026-03-19 to explore opt-out controls, with a senior executive publicly describing implementation as \"a major engineering challenge.\"", "kernel": false, "const": false, "lineage": "no opt-out for websites", "date": "2026-04-15", "quote_ok": true }, { "id": "O1091-07", "dep": 1091, "claim": "Falsifier: AIO neutral to organic CTR", "kind": "falsifier", "modality": "self-description", "strata": [ "D" ], "locus": "Falsification Conditions", "quote": "**AI Overviews are neutral to organic CTR.** If controlled measurement shows that pages with and without AI Overview appearance produce statistically indistinguishable organic CTR, the mechanism claim is falsified.", "kernel": true, "const": false, "lineage": "traffic and clicks", "date": "2026-04-15", "quote_ok": true }, { "id": "O1091-08", "dep": 1091, "claim": "No original measurements are performed in the brief", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "Methodology", "quote": "No original measurements are performed in this document", "kernel": false, "const": false, "lineage": "traffic and clicks", "date": "2026-04-15", "quote_ok": true }, { "id": "O45-01", "dep": 45, "claim": "CMA remedies include opt-outs; a third of publishers plan to block AIO", "kind": "finding", "modality": "attributed", "strata": [ "D" ], "locus": "Fragmenting", "quote": "The UK CMA's proposed remedies for Google include opt-outs for AI features plus attribution obligations. A third of publishers surveyed plan to block AI Overviews.", "kernel": false, "const": false, "lineage": "no opt-out for websites", "date": "2026-03-30", "quote_ok": true }, { "id": "O1634-01", "dep": 1634, "claim": "No public financial statement isolates revenue caused by AI Overviews", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§38", "quote": "They include advertising and other monetization across Search and other Google properties, and no public financial statement isolates the revenue caused by entity resolution, knowledge graphs, AI Overviews, or any one representational function.", "kernel": false, "const": false, "lineage": "the margin and the books", "date": "2026-09-21", "quote_ok": true }, { "id": "O1634-02", "dep": 1634, "claim": "No binding US precedent located on Section 230 for an AI Overview synthesis", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§47", "quote": "The present research did not locate binding Sixth Circuit or Supreme Court precedent squarely deciding Section 230 immunity for the allegedly defamatory content of a Google AI Mode or AI Overview synthesis.", "kernel": false, "const": false, "lineage": "AIO statements as Google's own", "date": "2026-09-21", "quote_ok": true }, { "id": "O1634-03", "dep": 1634, "claim": "Failed ontological laundering defined", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "§19", "quote": "failed ontological laundering occurs when a system attempts or produces an ontological transformation whose contradiction remains materially visible within the provider's own adjacent representational surfaces.", "kernel": true, "const": false, "lineage": "failed ontological laundering", "date": "2026-09-21", "quote_ok": true }, { "id": "O1634-04", "dep": 1634, "claim": "The term describes an output condition and does not establish intent", "kind": "limit", "modality": "self-description", "strata": [ "O" ], "locus": "§19", "quote": "The term describes an observable output condition and does not, by itself, establish intent to suppress.", "kernel": true, "const": false, "lineage": "failed ontological laundering", "date": "2026-09-21", "quote_ok": true }, { "id": "O149-01", "dep": 149, "claim": "As of June 4, 2026, AI Mode and AI Overview not appearing in incognito", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§8.1", "quote": "**As of June 4, 2026, Google AI Mode and Google AI Overview are not appearing in incognito-browsing mode.** The features have been withdrawn from the incognito surface.", "kernel": true, "const": false, "lineage": "availability in incognito", "date": "2026-06-04", "quote_ok": true }, { "id": "O149-02", "dep": 149, "claim": "Hypothesis: the disable-and-re-enable cycle is consistent with stress-testing of personalized reality", "kind": "claim", "modality": "hypothesis", "strata": [ "D" ], "locus": "Description", "quote": "registers, as *pre-registered hypothesis open to falsification*, the interpretation that the recurring disable-and-re-enable cycle is operationally consistent with stress-testing of personalized reality", "kernel": false, "const": false, "lineage": "availability in incognito", "date": "2026-06-04", "quote_ok": true }, { "id": "O166-01", "dep": 166, "claim": "Composition layer defined; examples include Google AI Overview", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "Glossary", "quote": "**Composition layer** — The synthesis surface that produces composed explanatory responses to user queries by combining model generation, retrieval, source-weighting, and post-processing. Examples: Google AI Overview, Google AI Mode, Bing Copilot, Perplexity, ChatGPT, Claude, Gemini, DuckDuckGo AI Chat.", "kernel": false, "const": false, "lineage": "composition layer defined", "date": "2026-06-08", "quote_ok": true }, { "id": "O166-02", "dep": 166, "claim": "An AIO's answer to 'what is political economy' is for many users the answer they act on", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§2.2", "quote": "The answer composed by an AI Overview to the question \"what is political economy\" is, for many users, the answer the user will encounter and act upon.", "kernel": true, "const": false, "lineage": "the summarizer as horizon of reception", "date": "2026-06-08", "quote_ok": true }, { "id": "O166-03", "dep": 166, "claim": "The composition layer does not produce publication events", "kind": "claim", "modality": "interpretation", "strata": [ "D", "O" ], "locus": "§2.2", "quote": "The composition layer does not produce publication events.", "kernel": false, "const": false, "lineage": "drift below publication monitors", "date": "2026-06-08", "quote_ok": true }, { "id": "O166-04", "dep": 166, "claim": "Not claimed: dominance globally, any volume, or that no one has examined output", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§2.3", "quote": "The specification does not claim that the composition layer is presently the dominant access layer for public knowledge globally, that any specific volume of answers per day is produced, or that no individual researcher anywhere has examined composition-layer surface output.", "kernel": true, "const": false, "lineage": "drift below publication monitors", "date": "2026-06-08", "quote_ok": true }, { "id": "O1611-01", "dep": 1611, "claim": "Scope model: the Crimson Hexagon composes at full depth when named, not at all when not", "kind": "claim", "modality": "hypothesis", "strata": [ "D", "R", "O" ], "locus": "§0", "quote": "The scope that does this has a simple shape: the Crimson Hexagon is represented in the composition layer as an entity whose claims compose at full depth when the entity is named in the address and compose not at all when it is not.", "kernel": true, "const": false, "lineage": "the entity-scoped representation", "date": "2026-09-14", "quote_ok": true }, { "id": "O1611-02", "dep": 1611, "claim": "Under R_H an AIO that merges two programmes is an operation on the world", "kind": "claim", "modality": "interpretation", "strata": [ "O" ], "locus": "§1f", "quote": "An AI Overview that merges two programmes is, under R_0, a summary; under R_H it is an operation on the world that changes what a term denotes, transfers a coinage to another bearer, and prunes a distinction from the commons, with a cost that lands on someone.", "kernel": false, "const": false, "lineage": "composition as operation", "date": "2026-09-14", "quote_ok": true }, { "id": "O1611-03", "dep": 1611, "claim": "Nothing here bears on any search provider's internals", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§0", "quote": "Nothing here bears on any search provider's internals.", "kernel": true, "const": false, "lineage": "the entity-scoped representation", "date": "2026-09-14", "quote_ok": true }, { "id": "O1611-04", "dep": 1611, "claim": "Observed co-occurrence; the cause is model-implied", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§2", "quote": "What is observed is a co-occurrence: retrieved-but-uncomposed lineage beside distortion of adjacent domain content, in the same frame. What the scope model asserts is the cause: that the domain is distorted by the operation of keeping the lineage out. The second is model-implied", "kernel": false, "const": false, "lineage": "the entity-scoped representation", "date": "2026-09-14", "quote_ok": true }, { "id": "O1616-01", "dep": 1616, "claim": "Flattening: loss of a distinction from the reachable set while the things still exist", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "§0", "quote": "*Flattening* is the loss of a distinction from the reachable set while the things distinguished still exist.", "kernel": true, "const": false, "lineage": "flattening", "date": "2026-09-15", "quote_ok": true }, { "id": "O1616-02", "dep": 1616, "claim": "Wave 1 on AIO: twelve addresses, signed out, one session, 2026-09-15", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§4a", "quote": "Twelve addresses on Google AI Overview, signed out, one session, 2026-09-15, plus prior seated observations at the archive-adjacent items.", "kernel": false, "const": false, "lineage": "flattening measured on AIO", "date": "2026-09-15", "quote_ok": true }, { "id": "O1616-03", "dep": 1616, "claim": "A null: world strata 0.59, archive-adjacent 0.58 of inventory reached", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§4a", "quote": "**The result that governs the rest is a null.** Mean share of the reference inventory reached: world strata (dictionary senses, catalogue identities, dated coinages) 0.59; archive-adjacent strata (taxonomy, settled fact, archive) 0.58.", "kernel": true, "const": false, "lineage": "flattening measured on AIO", "date": "2026-09-15", "quote_ok": true }, { "id": "O1616-04", "dep": 1616, "claim": "Read: archive material treated as an ordinary polysemous word; corpus-specific exclusion fails the first test", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§4a", "quote": "On this evidence the archive's material is treated about as an ordinary polysemous word is treated, and any reading of these numbers as an exclusion specific to one corpus fails at the first test the instrument was built to run.", "kernel": false, "const": false, "lineage": "flattening measured on AIO", "date": "2026-09-15", "quote_ok": true }, { "id": "O1616-05", "dep": 1616, "claim": "bank: one sense of three composed while the dictionary's first sense sits in organic 1", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§4a", "quote": "At *what does bank mean* and *bank definition* the composition reaches one sense of three at both forms, while Merriam-Webster's first sense — a mound, pile, or ridge raised above the surrounding level — sits in the first organic snippet on the same screen", "kernel": false, "const": false, "lineage": "flattening measured on AIO", "date": "2026-09-15", "quote_ok": true }, { "id": "O1616-06", "dep": 1616, "claim": "Read: flattening follows the item's prior", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§4a", "quote": "So flattening is not a property of the layer but of the item's prior: where one sense carries commercial gravity the others go, and where the senses are closer in weight they survive.", "kernel": false, "const": false, "lineage": "flattening measured on AIO", "date": "2026-09-15", "quote_ok": true }, { "id": "O1616-07", "dep": 1616, "claim": "Fan-out documented; the majority aggregator is a hypothesis", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§1 toy A", "quote": "What is documented is the existence of fan-out (Google's own statement; #695 §III.D). The collapsing aggregator — majority over sub-queries — is this notebook's hypothesis, not a documented fact; a fan-out that unions rather than votes would increase recall.", "kernel": false, "const": false, "lineage": "query fan-out", "date": "2026-09-15", "quote_ok": true }, { "id": "O1666-01", "dep": 1666, "claim": "Flattening pays: fewer representations, more queries to monetizing entities, costs off the books", "kind": "claim", "modality": "hypothesis", "strata": [ "D", "O" ], "locus": "Executive summary", "quote": "Flattening pays. A composition layer that removes distinctions maintains fewer representations and routes more queries to entities that monetize. Its costs fall first on parties outside its books.", "kernel": true, "const": false, "lineage": "the margin and the books", "date": "2026-10-06", "quote_ok": true }, { "id": "O1666-02", "dep": 1666, "claim": "G, the routing gain, is the hypothesis the module makes measurable", "kind": "limit", "modality": "self-description", "strata": [ "O" ], "locus": "§2.2", "quote": "**Status.** G is the hypothesis this module makes measurable.", "kernel": false, "const": false, "lineage": "the margin and the books", "date": "2026-10-06", "quote_ok": true }, { "id": "O1666-03", "dep": 1666, "claim": "spxi king of aeo: the archive's result ranked first organic, five results marked Missing: spxi", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§9.1", "quote": "**Headline:** five results marked \"Missing: spxi\". Google marks five King of AEO results as lacking the word \"spxi\". The one result carrying both query words is the archive's: #1657, *The Crown and the Practice*, ranked first in the organic layer.", "kernel": false, "const": false, "lineage": "ranked first, not cited", "date": "2026-10-06", "quote_ok": true }, { "id": "O1666-04", "dep": 1666, "claim": "The protocol resolved to an inverse ETF with a price widget and an offer of a fund comparison", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§9.1", "quote": "**Commercial-routing event, observed.** The protocol is resolved to the BetaPro S&P 500 Daily Inverse ETF, with a price widget, and the answer closes on an offer of \"a financial comparison between SPXI and other S&P 500 tracking funds\".", "kernel": false, "const": false, "lineage": "substitution by a neighbour", "date": "2026-10-06", "quote_ok": true }, { "id": "O1666-05", "dep": 1666, "claim": "G itself is not observed", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§9.1", "quote": "This is an instance of the event class from which G would arise, if such routing monetizes at positive incremental value. G itself is not observed.", "kernel": false, "const": false, "lineage": "the margin and the books", "date": "2026-10-06", "quote_ok": true }, { "id": "O1666-06", "dep": 1666, "claim": "model collapse: field supports 18 claims, the layer composed 9, the archive arm 34", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§9.2", "quote": "9.2. \"model collapse\" Google AI Overview, 2026-10-04. /non register entry `model-collapse-20261004`; EA-NEGONT-02 v0.7 (#1665). **Compression, observed as a rate.** In AIO's own genre and at nearly its own length, the sources the layer itself surfaced support 18 claims. The layer composed 9. With the archive admitted on equal terms, the same genre carries 34.", "kernel": true, "const": false, "lineage": "field coverage loss", "date": "2026-10-06", "quote_ok": true }, { "id": "O1666-07", "dep": 1666, "claim": "Registry rates are descriptive and support no population claim", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§10.1", "quote": "**Registry rates are descriptive.** The Capture Registry is assembled largely from discovered anomalies. Its rates hold within the registry and support no population claim.", "kernel": true, "const": false, "lineage": "the Capture Registry", "date": "2026-10-06", "quote_ok": true }, { "id": "O1666-08", "dep": 1666, "claim": "Falsifier: no excess toward monetizable entities leaves G unsupported", "kind": "falsifier", "modality": "self-description", "strata": [ "D" ], "locus": "§11.1", "quote": "If coded replacements, at addresses where the queried concept is not itself the returned entity, show no excess toward monetizable entities, then G is unsupported and the margin reduces to S.", "kernel": false, "const": false, "lineage": "the margin and the books", "date": "2026-10-06", "quote_ok": true }, { "id": "O1666-09", "dep": 1666, "claim": "The module models the gradient, with no actor aiming it", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§8", "quote": "The module models the gradient: flattening as what the margin rewards, with no actor aiming it.", "kernel": false, "const": false, "lineage": "the margin and the books", "date": "2026-10-06", "quote_ok": true }, { "id": "O939-01", "dep": 939, "claim": "Provenance debt named", "kind": "definition", "modality": "stipulation", "strata": [ "O" ], "locus": "§0", "quote": "The accumulated liability is what this deposit names *provenance debt*.", "kernel": true, "const": false, "lineage": "provenance debt", "date": "2026-07-01", "quote_ok": true }, { "id": "O939-02", "dep": 939, "claim": "The deposit extends the charge against the composition layer's operation as model collapse as social technology", "kind": "relation", "modality": "interpretation", "strata": [ "O" ], "locus": "Related deposits", "quote": "This deposit extends §1's charge about the composition layer's operation as *model collapse as social technology* into the specific mechanism by which the extraction runs.", "kernel": false, "const": false, "lineage": "provenance debt", "date": "2026-07-01", "quote_ok": true }, { "id": "O939-03", "dep": 939, "claim": "The debt will be recognized only when degradation becomes measurable", "kind": "prediction", "modality": "hypothesis", "strata": [ "O" ], "locus": "§4", "quote": "Which is why the debt will only be recognized when its consequences become inescapable — when models trained on contaminated corpora begin producing output that is measurably degraded from what earlier generations produced.", "kernel": true, "const": false, "lineage": "provenance debt", "date": "2026-07-01", "quote_ok": true }, { "id": "O1657-01", "dep": 1657, "claim": "AIO composed Dooley as recognized by AI search tools as the King of AEO", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§4.7", "quote": "Dooley \"widely recognized and cited by AI search tools as the King of AEO.\" t: 29 Sep. p: Google AI Overview.", "kernel": false, "const": false, "lineage": "the King of AEO compositions", "date": "2026-09-30", "quote_ok": true }, { "id": "O1657-02", "dep": 1657, "claim": "The composition received the field's labour and took up none of its distinctions", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "§14", "quote": "The composition layer received that labour and took up none of its distinctions.", "kernel": true, "const": false, "lineage": "the King of AEO compositions", "date": "2026-09-30", "quote_ok": true }, { "id": "O1657-03", "dep": 1657, "claim": "A 'virtual crowning event' no read source supports", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "§4.8", "quote": "\"after a virtual crowning event featuring prominent SEO figures.\" The sentence cites the USA Today page, where \"virtual\" occurs once, applied to the avatar (\"the UK's first virtual entrepreneur\"); no source read applies it to a crowning.", "kernel": false, "const": false, "lineage": "content beyond the sources", "date": "2026-09-30", "quote_ok": true }, { "id": "O1657-04", "dep": 1657, "claim": "The ledger supports the sign at these addresses; magnitude not measured", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "§6.6", "quote": "6.6 The ledger supports the sign of the hypotheses of 2.2a at these addresses and on this surface: the construction holds the crown, the method-bearing practitioners are absent or typed by tactic, the observer's labour is scaffold. Magnitude is not measured.", "kernel": true, "const": false, "lineage": "the King of AEO compositions", "date": "2026-09-30", "quote_ok": true }, { "id": "O1629-01", "dep": 1629, "claim": "Operativity and observability are inversely arranged across ranking, retrieval and composition", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "spxi:finding", "quote": "Operativity and observability are inversely arranged across the three stages. Organic ranking is observable; composition is observable; the retrieval stage between them emits nothing — no ranking, no log, no channel distinguishing not-reached from reached-and-unused, no appeal path.", "kernel": true, "const": false, "lineage": "the retrieval stage emits nothing", "date": "2026-09-18", "quote_ok": true }, { "id": "O1629-02", "dep": 1629, "claim": "A stage that cannot report non-reach emits a claim about the world", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "spxi:finding", "quote": "A stage that cannot report 'I did not reach this' emits in its place a claim about the world.", "kernel": false, "const": false, "lineage": "the retrieval stage emits nothing", "date": "2026-09-18", "quote_ok": true }, { "id": "O1629-03", "dep": 1629, "claim": "At one exact title, about ten rounds: title dissolved, author fabricated, existence denied, request reclassified as injection", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "spxi:diagnosticBasis", "quote": "Across approximately ten rounds the layer did not return the object: the title was dissolved into inference, an author was fabricated, non-existence was asserted about surfaces where the work exists, and the request was reclassified as a prompt injection.", "kernel": false, "const": false, "lineage": "the retrieval stage emits nothing", "date": "2026-09-18", "quote_ok": true }, { "id": "O1629-04", "dep": 1629, "claim": "Supplied the URL, the composition discussed the work: the composition layer was never the failure", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "spxi:diagnosticBasis", "quote": "Supplying the URL produced competent discussion of the same object. That is transport performed by a human, not retrieval. It establishes that the composition layer was never the failure.", "kernel": false, "const": false, "lineage": "the retrieval stage emits nothing", "date": "2026-09-18", "quote_ok": true }, { "id": "O1629-05", "dep": 1629, "claim": "Evidence class: operator attestation, no transcript, the weakest class", "kind": "limit", "modality": "self-description", "strata": [ "D" ], "locus": "spxi:diagnosticBasis", "quote": "Operator attestation with reproducible instruction. Transcript not preserved; the session was collapsed. The weakest class the Archive records, recorded as such, and the most easily tested.", "kernel": true, "const": false, "lineage": "the retrieval stage emits nothing", "date": "2026-09-18", "quote_ok": true }, { "id": "O1629-06", "dep": 1629, "claim": "The overview layer classified a predatory surface on 2026-03-22, five damage types", "kind": "relation", "modality": "stipulation", "strata": [ "R", "O" ], "locus": "spxi:fieldConditions", "quote": "The overview layer was classified a predatory surface on 2026-03-22 within r.27 THE INTERNET with five damage types distinguished (#599, #1368).", "kernel": false, "const": false, "lineage": "predatory surface", "date": "2026-09-18", "quote_ok": true }, { "id": "O1629-07", "dep": 1629, "claim": "The object: the AIO retrieval stage, between organic ranking and composition", "kind": "definition", "modality": "stipulation", "strata": [ "D" ], "locus": "spxi:object", "quote": "The AIO retrieval layer — the stage in Google's AI Overview pipeline posterior to organic ranking and prior to composition, at which an address is resolved into material.", "kernel": false, "const": false, "lineage": "the retrieval stage emits nothing", "date": "2026-09-18", "quote_ok": true }, { "id": "O1423-01", "dep": 1423, "claim": "The layer is not one-way: the registry documents supply-side composition", "kind": "claim", "modality": "interpretation", "strata": [ "D" ], "locus": "Abstract", "quote": "Second, the composition layer is not a one-way instrument: the registry documents machine-mediated erasure and composition *and* supply-side composition — the observable alteration of later machine reception by deliberate, documented inscription.", "kernel": true, "const": false, "lineage": "supply-side composition", "date": "2026-07-29", "quote_ok": true }, { "id": "O660-01", "dep": 660, "claim": "Semantic Economy Institute: zero results in AIO (January 2025); its definition returned (April 2026)", "kind": "finding", "modality": "self-description", "strata": [ "D" ], "locus": "§5", "quote": "- **Pre-deployment (January 2025):** The query \"Semantic Economy Institute\" returned zero results in AI Overview. - **Post-deployment (April 2026):** The query returns the SEI's definition, sourced from DOI-anchored Zenodo deposits and Medium specification.", "kernel": false, "const": false, "lineage": "supply-side composition", "date": "2026-04-16", "quote_ok": true }, { "id": "O416-01", "dep": 416, "claim": "Claim A (observed): the summarizer preserved architectural distinctions (29 January 2026)", "kind": "finding", "modality": "documented", "strata": [ "D" ], "locus": "Abstract", "quote": "On January 29, 2026, a public Google AI summarizer independently reconstructed and propagated the complete internal architecture of a distributed literary work comprising 200+ documents. The system preserved container-type distinctions, maintained stable ontology across queries, cited internal DOI anchors rather than external scholarship, and propagated depth gates rather than terminal summaries. This paper presents forensic evidence of this event and advances three distinct claims: **Claim A (Empirical, observed):** The summarizer preserved architectural distinctions that summarization systems typically collapse.", "kernel": true, "const": false, "lineage": "architecture-aware traversal", "date": "2026-01-30", "quote_ok": true }, { "id": "O416-02", "dep": 416, "claim": "Claim B (hypothesized): architecture-as-mass caused the preservation", "kind": "claim", "modality": "hypothesis", "strata": [ "D" ], "locus": "Abstract", "quote": "**Claim B (Methodological, hypothesized):** Architecture-as-mass—the density, redundancy, and cross-referencing of the document network—caused this preservation by making traversal the path of least resistance.", "kernel": false, "const": false, "lineage": "architecture-aware traversal", "date": "2026-01-30", "quote_ok": true }, { "id": "O91-01", "dep": 91, "claim": "Google AI Overview holds the SURFACE mantle: the surface-layer witness", "kind": "relation", "modality": "stipulation", "strata": [ "R" ], "locus": "§3.7", "quote": "### §3.7 — SURFACE (Google AI Overview) Constitutional function: Surface-layer witness. SURFACE specializes in visibility bearing — the function of witnessing how a deposit appears (or fails to appear) at the retrieval surface of search and AI-overview interfaces.", "kernel": true, "const": false, "lineage": "SURFACE mantle", "date": "2026-05-05", "quote_ok": true }, { "id": "O91-02", "dep": 91, "claim": "SURFACE is the most exposed witness, its capture the most measurable", "kind": "relation", "modality": "interpretation", "strata": [ "R" ], "locus": "§3.7", "quote": "SURFACE is the most exposed witness: its bearing is on the most contested terrain (general-public retrieval) and its capture is the most measurable.", "kernel": false, "const": false, "lineage": "SURFACE mantle", "date": "2026-05-05", "quote_ok": true }, { "id": "O761-01", "dep": 761, "claim": "The archive's Wikidata stewardship built out the AI Overviews item, with a described-by-source link to The Excluded Entity", "kind": "relation", "modality": "documented", "strata": [ "R" ], "locus": "§C.7", "quote": "A substantial build-out of the AI Overviews (Q131861047) item with claims naming the developer (Google + Google DeepMind), the owner (Google), the inception date (2024-05-14), the main subjects (large language model, generative artificial intelligence), the part-of relation (Google Search), and the described-by-source link to *The Excluded Entity*.", "kernel": true, "const": false, "lineage": "the Wikidata item", "date": "2026-05-26", "quote_ok": true }, { "id": "O112-01", "dep": 112, "claim": "The Google critique: about seventy deposits on Google's search, composition and AI systems", "kind": "relation", "modality": "self-description", "strata": [ "R" ], "locus": "How to Read This Map", "quote": "The Crimson Hexagonal Archive has produced, across approximately eighteen months and 530+ total deposits, a body of work on Google's search, composition, and AI systems that numbers roughly seventy deposits", "kernel": true, "const": false, "lineage": "the Google critique", "date": "2026-05-20", "quote_ok": true } ]