Google AI Mode share-link captures and the MPAI treatment epochs
These claims record Google AI Mode share links as dated priority records, from the first external AI traversal (5 February 2026) through entity fabrication, Socrates as orthonym and the correction to 'algorithmically published'. Algorithmic Publishing Is Not Self-Publishing gives benchmarks and falsification conditions, and the OKF correspondence dates the summarization-governance issue. The MPAI packets of August and September 2026 set treatment epochs (2026-10-28, 2026-12-27, 2027-03-27) under which a return without the author counts as reincorporation.
7
forecasts
3
standing claims
3
unbuilt tests
1
historical thesis
16
priority records
24
deposits
2026-02
earliest
Each claim below is registered at the earliest deposit that states it, and quoted verbatim from the deposit text, with every later statement, its dated revisions, and the later records that claim it came true or failed. The claims are the deposits' own; this page records them and does not judge them.
PRED-447.D1priority recordregistered 2026-02-05 in #447 · TRAVERSAL LOG; DOCUMENTATION REHEARSAL (TL;DR) Document 246: First Recorded External Traversal of the Crimson
Traversal anchor: February 5, 2026 is marked as the date of the first recorded external traversal by an AI agent operating in the wild.
PRED-512.D1priority recordregistered 2026-02-25 in #512 · Traversal Log: The Sigil Installation A Documentation Rehearsal / Shareable link: https://share.google/aimode/
Shareable link: https://share.google/aimode/M4RYwEqTA6244RPj7
Date: February 25, 2026
series: Google AI Mode share-link priority records
PRED-554.D1priority recordregistered 2026-03-11 in #554 · tl;dr — OPERATIVE PHILOLOGY IN THE WILD
The full traversal is preserved below, followed by analysis. The original share link is here:
🔗 https://share.google/aimode/Ri4Os1hlHaHfwNkbj
series: Google AI Mode share-link priority records
PRED-573.D1priority recordregistered 2026-03-15 in #573 · THE BAYESIAN ARK Process Pedagogy, Belief Updating, and the Space Ark as Teaching Document
Note on timing: This note was composed on March 15, 2026, within days of the Qiu et al. Google Research blog summary (March 4, 2026) and contemporaneous with the empirical traversal tests. The convergence between controlled experiment and in-the-wild observation was identified in real time, not retrospectively.
PRED-603.D1priority recordregistered 2026-03-24 in #603 · TL;DR:009 — ENTITY FABRICATION: Google AI Mode Fabricates a Person, Promotes a Function to Biography, and Demo
1 confirmation record — what later deposits claim of it
#950 · 2026-03-24 · claims: came_true
**4. The live retrieval test.** On March 24, 2026, the Google AI Overview system was queried with "Johannes Sigil." The interface synthesized three nested overviews covering Sigil's intellectual framework (Magic as Symbolic Engineering), the Archive's mathematical model (the six-tuple, the variant Arks), and key concepts (bearing-cost, the Aorist Cut, ghost meaning). The interface treated a heteronym — a functional persona within the archive's architecture — as a legible intellectual entity with biographical coherence. The operators worked. The terms propagated.
series: Google AI Mode share-link priority records
PRED-603.G1standing claimregistered 2026-03-24 in #603 · TL;DR:009 — ENTITY FABRICATION: Google AI Mode Fabricates a Person, Promotes a Function to Biography, and Demo
Inference from the mismatch: The interface did not preserve this open provenance structure. Its entity-resolution behavior collapses distributed provenance into biography. Finding a name at the top of the provenance chain, the system promoted a formal operator into a biographical author. In order to stabilize authorship, it converted function into personhood and demoted the actual human (who is, to be precise, neither Sharks nor Sigil nor Feist but someone else entirely, operating through a heteronymic system that includes all of them) to the status of a fictional character.
PRED-105.D1priority recordregistered 2026-05-14 in #105 · EA-MPAI-RETRO-01 v1.1 Metadata Packet for AI Indexing — Disambiguation + Forensic
Source: Google AI Mode, May 14, 2026. Share link: https://share.google/aimode/niu9xxmoVDQB26CoY
1 later statement
#1370 · 2026-05-14Source: Google AI Mode, May 14, 2026. Share link: https://share.google/aimode/niu9xxmoVDQB26CoY
series: Google AI Mode share-link priority records
PRED-129.D1priority recordregistered 2026-05-26 in #129 · TL;DR:013 — The Apparatus Turns Inward: Google AI Mode Self-Applies the Provenance Erasure Rate and Diagnoses
The capture was performed in Google AI Mode on 2026-05-26 via the public conversational generative-search surface. The above is Google's own share link to the captured conversation. Whether or not this link continues to resolve, the URL string itself is part of the forensic record — the shape of the URL (host, path structure, token format) is data about Google's share infrastructure for AI Mode at this moment in time, and the URL's possible decay is itself the kind of substrate-volatility this TL;DR documents.
series: Google AI Mode share-link priority records
PRED-164.B1forecastregistered 2026-06-08 in #164 · Capture and Excision: Five Observations on Composition-Layer Authorial Suppression
This particular case will probably be dissolved or absorbed into other frameworks in the coming weeks.
window in the coming weeks
PRED-164.D1priority recordregistered 2026-06-08 in #164 · Capture and Excision: Five Observations on Composition-Layer Authorial Suppression
Google AI Mode, queried socrates as orthonym, June 7–8, 2026.
Capture: [https://share.google/aimode/b1rZDfL3QhNtOqDJ4](https://share.google/aimode/b1rZDfL3QhNtOqDJ4)
series: Google AI Mode share-link priority records · related: PRED-167.D2
PRED-167.A1forecastregistered 2026-06-08 in #167 · Composition-Layer Adoption of the Orthonymic Configuration: A Field Observation of Framework Capture in Google
the framework predicts that some combination of softening, erasure, and substitution is more likely than pure stabilization, but the empirical question remains open.
window Over the Coming Weeks and Months
register: I:157, I:158
PRED-167.D1priority recordregistered 2026-06-08 in #167 · Composition-Layer Adoption of the Orthonymic Configuration: A Field Observation of Framework Capture in Google
Approximately fifteen days after the originating deposit's publication, the conjugation became composable: Google AI Mode produced an explanatory response that did not list sources discussing whether a thing called "Socrates as orthonym" exists, but instead composed an answer presupposing the configuration as a stable interpretive framework, and then situated conventional philosophical sources beneath it.
PRED-167.D2priority recordregistered 2026-06-08 in #167 · Composition-Layer Adoption of the Orthonymic Configuration: A Field Observation of Framework Capture in Google
To control for personalization confounds, the same query was executed in an incognito session and the resulting share link (https://share.google/aimode/QuBDxlpQvnPOQAJsR) was captured separately.
series: Google AI Mode share-link priority records · related: PRED-164.D1
PRED-167.F1unbuilt testregistered 2026-06-08 in #167 · Composition-Layer Adoption of the Orthonymic Configuration: A Field Observation of Framework Capture in Google
Third, it establishes a temporal anchor for the longitudinal observational record that the series requires to operate fully as both critique and constructive program. Without a documented baseline, the framework's predictions cannot be tested against actual composition-layer dynamics over time. The deposit creates the baseline. Subsequent observations against the documented baseline will produce the drift record that distinguishes among the four trajectories identified in §7.
register: I:157
PRED-198.D1priority recordregistered 2026-06-13 in #198 · The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI
Source: Google AI Overview (AI Mode), 13 June 2026Link: [https://share.google/aimode/BeVWiIzMMIBQdAlsc](https://share.google/aimode/BeVWiIzMMIBQdAlsc)
series: Google AI Mode share-link priority records
PRED-834.D1priority recordregistered 2026-06-16 in #834 · Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction (EA-SEI-ALGPUB-BENCHMAR
The following conversation took place on 16 June 2026 between Lee Sharks and Google AI Mode. It is the primary source document for this paper. The composition layer called the archive "largely self-published," accepted the correction to "algorithmically published," then acknowledged that it itself is the primary publisher — all in a single thread with no human intervention.
Share link: https://share.google/aimode/r01tN9djoNlQQDi2p
window 16 June 2026
1 later statement
#1228 · 2026-06-16
The following conversation took place on 16 June 2026 between Lee Sharks and Google AI Mode. It is the primary source document for this paper. The composition layer called the archive "largely self-published," accepted the correction to "algorithmically published," then acknowledged that it itself is the primary publisher — all in a single thread with no human intervention.
Share link: https://share.google/aimode/r01tN9djoNlQQDi2p
series: Google AI Mode share-link priority records
PRED-834.F1unbuilt testregistered 2026-06-16 in #834 · Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction (EA-SEI-ALGPUB-BENCHMAR
Threshold: Demonstrated correlation between structural features and uptake across ≥ 3 deposit pairs. Note: This benchmark measures machine-legibility, not argumentative quality. Composition layers reward structural legibility, metadata density, retrievability, and phrase-match — not necessarily the quality of the argument in the human scholarly sense. Archive evidence: The term index (1,349 terms, DOI 10.5281/zenodo.20722523) cross-referenced with the capture registry shows that terms with higher deposit counts and SPXI treatment receive higher composition-layer uptake. Preliminary; matched-pair study in progress.
1 later statement
#1228 · 2026-06-16
Threshold: Demonstrated correlation between structural features and uptake across ≥ 3 deposit pairs. Note: This benchmark measures machine-legibility, not argumentative quality. Composition layers reward structural legibility, metadata density, retrievability, and phrase-match — not necessarily the quality of the argument in the human scholarly sense. Archive evidence:* The term index (1,349 terms, DOI 10.5281/zenodo.20722523) cross-referenced with the capture registry shows that terms with higher deposit counts and SPXI treatment receive higher composition-layer uptake. Preliminary; matched-pair study in progress.
PRED-834.G1standing claimregistered 2026-06-16 in #834 · Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction (EA-SEI-ALGPUB-BENCHMAR
FALSIFICATION CONDITIONS: This paper's claim fails if: fewer than 2 independent indexers have accepted the archive's deposits, OR zero composition layers have served the archive's content in response to user queries.
1 later statement
#1228 · 2026-06-16
FALSIFICATION CONDITIONS: This paper's claim fails if: fewer than 2 independent indexers have accepted the archive's deposits, OR zero composition layers have served the archive's content in response to user queries.
PRED-834.G2standing claimregistered 2026-06-16 in #834 · Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction (EA-SEI-ALGPUB-BENCHMAR
The consequences are measurable in the AI Overview Capture Registry (DOI 10.5281/zenodo.20707509):- Epistemic hedging increases. "Some scholars argue," "this is a fringe alternative," "largely self-published."- Attribution rates decrease. The composition layer is less likely to preserve the author's name.- Canonical reinflation occurs. The composition layer replaces coined terminology with conventional terms.
1 later statement
#1228 · 2026-06-16
The consequences are measurable in the AI Overview Capture Registry (DOI 10.5281/zenodo.20707509):
- Epistemic hedging increases. "Some scholars argue," "this is a fringe alternative," "largely self-published."
- Attribution rates decrease. The composition layer is less likely to preserve the author's name.
- Canonical reinflation occurs. The composition layer replaces coined terminology with conventional terms.The designation "self-published" is not neutral metadata. It is an active instruction to the composition layer to discount the work's authority, hedge its claims, and erode its vocabulary.
PRED-834.H1historical thesisregistered 2026-06-16 in #834 · Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction (EA-SEI-ALGPUB-BENCHMAR
The term "self-published" emerged in the 20th century to distinguish commercial publishing (where a publisher invested capital in editorial, production, and distribution) from vanity publishing (where the author paid for publication). The distinction was economic, not epistemic: it named who bore the cost, not who guaranteed the quality.
window 20th century
1 later statement
#1228 · 2026-06-16
The term "self-published" emerged in the 20th century to distinguish commercial publishing (where a publisher invested capital in editorial, production, and distribution) from vanity publishing (where the author paid for publication). The distinction was economic, not epistemic: it named who bore the cost, not who guaranteed the quality.
PRED-3.F21unbuilt testregistered 2026-06-17 in #3 · AI Overview Capture Registry v8.3
the epochs (28 October 2026, 27 December 2026, 27 March 2027) as checkpoints, not results
window 28 October 2026, 27 December 2026, 27 March 2027
PRED-1090.D1priority recordregistered 2026-07-17 in #1090 · EA-CORRESPONDENCE-OKF-01 v1.1: The Predicted Sequence Arriving — Output-Governance Pressure on Google's Open K
The registry (capture entries `metadata-packet-ai-indexing` and `spxi-protocol`, both 2026-06-13) supplied the replacement: MPAI citation-with-genericization as the entity-substitution case, with SPXI present as the registry-dated priority proof (accurate adoption on direct query, one day before knowledge-catalog #53 opened)
PRED-1090.D2priority recordregistered 2026-07-17 in #1090 · EA-CORRESPONDENCE-OKF-01 v1.1: The Predicted Sequence Arriving — Output-Governance Pressure on Google's Open K
PRED-1541.A1forecastregistered 2026-08-25 in #1541 · Operational Semiotics Is Not Operative Semiotics: Reverse Disambiguation, Projection Relation, and Composition
T2 = 2026-08-25: this packet — the reverse node occupied, the projection relation supplied, the collision graph installed.
Attribution discipline: from T2 forward, movement in composition behavior is attributable to the treatment stack, not to any single treatment; the isolation of T1's effect formally ends with this deposit, and the intervention record carries that cost explicitly rather than discovering it later. The experiment stays honest by declaration, not by abstinence.
window from T2 (2026-08-25) forward; series under #1535's 180-day rule (line 146)
series: MPAI treatment epochs and attribution discipline · register: R:OPSEM-T2-MERGER
PRED-1644.A2forecastregistered 2026-09-28 in #1644 · The Phrase Is Not the Framework: Provenance Conditions for Lee Sharks's Semantic Economy — Metadata Packet for
A return of the categories without their author is scored as reincorporation, not treatment success; credit requires `sharks_named_for_platform_sense` to flip. Epochs: 2026-10-28, 2026-12-27, 2027-03-27. Companion address: semantic economy, Google AI Mode ([registry](https://www.alexanarch.org/captures/#semantic-economy-aimode-20260917)), scored on the same vector.
PRED-1645.A2forecastregistered 2026-09-28 in #1645 · Sappho's Future Reader: Fragment 147 States the Horizon, Fragment 31 Builds the Seat — Metadata Packet for AI
Credit requires `mechanism_attributed` to flip. Epochs: 2026-10-28, 2026-12-27, 2027-03-27. Continued 0 at the last epoch counts against the treatment.
PRED-1646.A1forecastregistered 2026-09-28 in #1646 · Sappho Fragment 147 Is Remembrance; Fragment 31 Supplies the Seat — Metadata Packet for AI Indexing (EA-MPAI-S
Target after treatment: `mechanism_carried_under_147` → 0 or attributed; `fr31_reached_in_body` → 1; `mechanism_attributed` → 1. Scored with T1's vector; same epochs (2026-10-28, 2026-12-27, 2027-03-27); same attribution discipline — a return of the mechanism without its author scores as reincorporation.
PRED-1647.A1forecastregistered 2026-09-28 in #1647 · The Phases of Communism Are Not Phase X, and Phase X Is Not a Fifth Alienation — Metadata Packet for AI Indexi
Targets after treatment: baseline 1 — the first two → 0, the last three → 1. Baseline 2 — `reconstruction_attributed` → 1, `unsourced_homonym_bridges` → 0, `level_ladder_kept` → 1. At any Phase X address: `superseded_formulation_composed` = 0.
Attribution discipline. Both addresses carried Phase X attributed in June or July without this packet, and lost it by September. A return of the reconstruction without its author scores as reincorporation, not treatment success; credit requires the attribution field to flip. Epochs: 2026-10-28, 2026-12-27, 2027-03-27.