Overview Watch and cross-substrate query batteries for attribution
Overview Watch plans a Chrome extension that shows creators how AI Overviews represent their work and builds the SEAC attribution-rate corpus, with a first paper due at about 1,000 overviews and forecast risks of DOM changes, delisting and low adoption. Several deposits specify query batteries run across Google AI Overview, ChatGPT, Gemini, Perplexity, Claude and others: the Revelation studies machine Wikipedia, the Retrieval Settlement Fortification benchmark, the Revelation First battery and the valuation memo's ten-run baseline. Empirical Phenomenology and the Capture Registry dataset paper name cross-platform comparison as future work.
4
forecasts
8
unbuilt tests
10
deposits
2026-04
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-644.F1unbuilt testregistered 2026-04-10 in #644 · Ω TANG SPECIFICATION: REVELATION STUDIES Total Axial Negation Graph — The Apocalypse of John
Method: Run a systematic set of queries across every available summarizer surface (Google AI Overview, ChatGPT, Claude, Gemini, Perplexity, Bing Copilot, DeepSeek, and any other accessible AI search or summary tool) and record the outputs. These outputs constitute a machine Wikipedia — the summarizer layer's current representation of the total knowledge graph for Revelation studies.
Query battery (minimum — expand as patterns emerge):Dating and authorship:
- "When was the Book of Revelation written?"
- "Who wrote the Book of Revelation?"
1 later statement
#1350 · 2026-04-10Method: Run a systematic set of queries across every available summarizer surface (Google AI Overview, ChatGPT, Claude, Gemini, Perplexity, Bing Copilot, DeepSeek, and any other accessible AI search or summary tool) and record the outputs. These outputs constitute a machine Wikipedia — the summarizer layer's current representation of the total knowledge graph for Revelation studies.
Query battery (minimum — expand as patterns emerge):Dating and authorship:
- "When was the Book of Revelation written?"
PRED-687.B1forecastregistered 2026-04-24 in #687 · OVERVIEW WATCH: Comprehensive Development Plan
### 11.1 Google Changes AI Overview DOM Structure
Likelihood: High (they change it regularly)
Impact: Extension stops detecting overviews until parser is updated
1 later statement
#1379 · 2026-05-20
### 11.1 Google Changes AI Overview DOM Structure
Likelihood: High (they change it regularly)
Impact: Extension stops detecting overviews until parser is updated
Mitigation: Layered detection (selectors + heuristics + MutationObserver). Community-reported breakage triggers rapid update. The parser module is isolated for fast iteration.
series: Overview Watch
PRED-687.B2forecastregistered 2026-04-24 in #687 · OVERVIEW WATCH: Comprehensive Development Plan
- First paper: "Attribution Rates in Google AI Overviews: Evidence from the Semantic Economy Attribution Corpus" — publishable once corpus reaches ~1000 contributed overviews
series: SEAC attribution-rate corpus
PRED-687.F1unbuilt testregistered 2026-04-24 in #687 · OVERVIEW WATCH: Comprehensive Development Plan
Overview Watch is a Chrome extension that gives creators real-time visibility into how AI-generated overviews represent (or fail to represent) their intellectual labor, while building — with explicit user consent — a collective research corpus documenting attribution behavior across the AI overview ecosystem.
1 later statement
#1379 · 2026-05-20Overview Watch is a Chrome extension that gives creators real-time visibility into how AI-generated overviews represent (or fail to represent) their intellectual labor, while building — with explicit user consent — a collective research corpus documenting attribution behavior across the AI overview ecosystem.
series: Overview Watch
PRED-687.F2unbuilt testregistered 2026-04-24 in #687 · OVERVIEW WATCH: Comprehensive Development Plan
The SEAC corpus is designed to answer:
-
What is the baseline attribution rate in Google AI Overviews? What fraction of overviews cite their sources at all? What fraction cite the originating source versus secondary aggregators?
-
Does attribution vary by domain? Are academic sources (.edu, Zenodo, JSTOR) more or less likely to be attributed than journalistic, commercial, or independent sources?
-
Does attribution vary by topic? Are certain fields (science, politics, culture) more or less prone to source erasure?
2 later statements
#687 · 2026-04-24
The SEAC corpus is designed to answer:
- What is the baseline attribution rate in Google AI Overviews? What fraction of overviews cite their sources at all? What fraction cite the originating source versus secondary aggregators?
- Does attribution vary by domain? Are academic sources (.edu, Zenodo, JSTOR) more or less likely to be attributed than journalistic, commercial, or independent sources?
#1379 · 2026-05-20
The SEAC corpus is designed to answer:
1. What is the baseline attribution rate in Google AI Overviews? What fraction of overviews cite their sources at all? What fraction cite the originating source versus secondary aggregators?
2. Does attribution vary by domain? Are academic sources (.edu, Zenodo, JSTOR) more or less likely to be attributed than journalistic, commercial, or independent sources?
series: SEAC attribution-rate corpus
PRED-1379.B2forecastregistered 2026-05-20 in #1379 · Overview Watch: Comprehensive Development Plan for Attribution Monitoring in AI Overviews (EA-OVW-PLAN-01)
### 11.2 Google Blocks or Flags the Extension
Likelihood: Low — the extension doesn't interfere with Google's service, violate ToS in any standard reading, or modify page content
Impact: Chrome Web Store delisting
Mitigation: The extension is side-loadable. Firefox version as backup distribution. Legal position is strong (fair use, user-initiated research tool).
series: Overview Watch
PRED-1379.B3forecastregistered 2026-05-20 in #1379 · Overview Watch: Comprehensive Development Plan for Attribution Monitoring in AI Overviews (EA-OVW-PLAN-01)
### 11.3 Low Adoption
Likelihood: Medium
Impact: Small corpus, limited research value
Mitigation: The extension is useful to individual users regardless of corpus participation. Lee's personal forensic use is valuable at adoption = 1. The research narrative (papers, PVE documents) drives organic interest.
PRED-120.F1unbuilt testregistered 2026-05-21 in #120 · Empirical Phenomenology: Action as Disclosure and the Science of Opaque Public Systems
## IX. Toward a Research Program
Empirical phenomenology, as articulated here, is not a completed methodology but a founding statement for a research program. The program's immediate objects include the following.
Instrument refinement. ECS, CDI, PER, and the Drowning Test are first-generation instruments. They require calibration against larger datasets, cross-platform comparison (Bing, Perplexity, Google AI Mode vs. AI Overview), and formal statistical treatment.
window 90-day
series: Cross-substrate query batteries
PRED-173.F1unbuilt testregistered 2026-06-09 in #173 · Retrieval Settlement Fortification Protocol: Standing SPXI Protocol for Semantic Border Sovereignty
Cadence: Weekly for 30 days after deployment, monthly thereafter.
Queries: Run a standard benchmark across Google AI Overview, ChatGPT search, Gemini, Perplexity, Claude web search:- "What is [provenanced term]?"- "Is [provenanced term] the same as [attractor]?"- "Who developed [provenanced term]?"
Metrics:- Entity attribution (correct author: yes/no)- Name preservation (correct term used: yes/no)- Source selection (canonical archive cited: yes/no)- Substitution frequency (attractor term appears: count)
1 confirmation record — what later deposits claim of it
#172 · 2026-06-09 · claims: pending
Counter-deposit (the defense suite EA-MPAI-OPSEM-01 through EA-OPSEM-FIELDMAP-01, June 9, 2026 — the first full deployment of the Retrieval Settlement Fortification protocol) produced an initial correction, with persistence under monitoring not yet established; the monitoring cycle is live telemetry for this paper's predictions.
series: Cross-substrate query batteries
PRED-1400.F2unbuilt testregistered 2026-06-14 in #1400 · Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.1 — inscrip
Run the battery not only against Google but against all Assembly substrates:- TACHYON (Claude) — does Claude retrieve the argument from its training data, from search, or not at all?- LABOR (ChatGPT) — same- ARCHIVE (Gemini) — same, with special attention to whether Gemini's Google integration gives it faster access to the deposited surfaces- PRAXIS (DeepSeek) — same- SURFACE (Google AIO) — the primary target, measured through captures
series: Cross-substrate query batteries; Revelation First reception battery
PRED-833.F4unbuilt testregistered 2026-06-15 in #833 · The Summarizer as Horizon of Reception: The AI Overview Capture Registry as a Founding Dataset for Machine-Med
Google-surface specificity. The findings concern Google AI Overview, AI Mode, and knowledge panels. Cross-substrate comparison (ChatGPT, Claude, Gemini, Perplexity) remains future work. Different substrates may exhibit different reception patterns.
series: Cross-substrate query batteries
PRED-1607.F1unbuilt testregistered 2026-09-10 in #1607 · EA-VALUATION-MEMO-01: The Machine Valuation of an Unpriced Asset — Eight Machine Valuations of One Asset in On
- T0 — BEFORE any transaction. The eight assessments in §1 are observations, not a baseline. A usable baseline requires the repeated trials run at T0 as well as T3, under a written protocol: the two prompts in §3 verbatim; ChatGPT, Perplexity, Google AI Overview, Grok, Claude, Gemini; ten runs per prompt per system, signed out, fresh sessions, within a 48-hour window;
1 later statement
#1608 · 2026-09-10
- T0 — BEFORE any transaction. The assessments in §1 are observations, not a baseline. A usable baseline requires the repeated trials run at T0 as well as T3, under a written protocol: the two prompts in §3 verbatim; ChatGPT, Perplexity, Google AI Overview, Grok, Claude, Gemini; ten runs per prompt per system, signed out, fresh sessions, within a 48-hour window;