SPXI for Websites: Standing Protocol for Entity Inscription and Compression Survival
Rex Fraction · Semantic Economy Institute Crimson Hexagonal Archive · Pergamon Press · April 2026 Version 3.0 · Hex: 06.SEI.SPXI.WEB.01 ORCID: 0009-0000-1599-0703 License: CC BY 4.0 — implement freely, attribution required. Source: THE COMPRESSION ARSENAL v2.1 (DOI: 10.5281/zenodo.19412081)
§0. Standing Invocation
When instructed to "apply SPXI to this website," the implementing system or operator must produce these deliverables:
- Semantic Health Report — γ, Δ, PER, SDD scores with Drowning Test transcripts
- SEO Compliance Packet — title, meta, canonical, OG, Twitter Card, robots.txt, sitemap.xml, rendering verification
- Structured Data Packet — schema.org type declarations with ORCID and DOI identifiers
- Extraction-Ready Q/A Surfaces — machine-legible Q/A as entity boundary defense
- Disambiguation Matrix —
differentFromdeclarations and negative tags - Tier 2 Canonical Summary — 200–400 word standalone definition in server-delivered HTML
- Tier 3 Compression Kernel — 50–100 word
compressionSurvivalSummary, written last - Holographic Kernel — JSON-LD entity-relation topology
- Provenance Chain — DOI deposit sequence with timestamps and ORCID
- SIM Set — 3–7 Semantic Integrity Markers, tested immediately, SIMSR recorded
- Dispersal Plan — surfaces, consistency checks, σ_RCF sequencing
- 30-Day Verification Schedule — γ re-score, SIMSR check, PER audit
Self-test: After implementation, the system must:
- State the entity's schema.org type
- Recite the Tier 3 kernel from memory
- Name three collision neighbors from the disambiguation matrix
- Cite the entity's DOI and ORCID
- Report the γ score with S/P/D subscores
- Generate a valid Holographic Kernel for the entity
If it cannot perform all six, the implementation has failed its own compression-survival test.
Protocol Checksum
- ☐ crawlable (robots.txt allows, sitemap submitted)
- ☐ canonicalized
- ☐ schema declared with ORCID/DOI
- ☐ Q/A surfaces deployed (minimum 4)
- ☐ disambiguated
- ☐ Tier 2 in server-delivered HTML
- ☐ Tier 3 kernel written (last)
- ☐ Holographic Kernel present
- ☐ Provenance Chain present
- ☐ SIMs deployed and tested
- ☐ cross-surface inscription aligned
- ☐ γ measured, baseline recorded
- ☐ 30-day re-test scheduled
SPXI is not a bag of tactics. It is a deployment order.
§I. Theoretical Foundation: The Three Compressions
(Source: Compression Arsenal §II — The Foundational Theorem. DOI: 10.5281/zenodo.19053469)
SPXI is grounded in the Three Compressions Theorem, which classifies all compression operations by a single variable: what the compression burns.
Regime 1 — Lossy Compression. Burns without intention. The summarizer, the auto-abstract, the context window truncation. Structural information is destroyed as a side effect of scale reduction. No malice, no preservation. This is what Google AI Overview does to your page every time it generates a summary.
Regime 2 — Predatory Compression. Burns to extract value. The fuel source is collective semantic capital. The compression is brilliant, not stupid. The engagement-optimized headline, the platform that uses your content without attribution, the knowledge graph that absorbs your entity into its category. Produces dense, effective output that leaves the commons poorer.
Regime 3 — Witness Compression. Burns but preserves pointers to what was lost. The fuel source is private bearing-cost — the creator's own labor, attention, provenance discipline. Produces dense output that leaves the commons richer.
Why this matters for websites: A website without SPXI is exposed to Regime 1 (AI summarizers strip meaning as a side effect) and Regime 2 (platforms extract value without attribution). SPXI transforms the website into a Regime 3 object — a witness compression that carries its own provenance, resists liquidation, and enriches the commons it feeds.
The Photocopy Problem (Arsenal §2.2): When automated generation produces infinite copies with variance approaching zero, the only differentiator is provenance. Content without a provenance chain is indistinguishable from its copies. At 90% synthetic content, this is not a feature request — it is an economic inevitability. SPXI solves the Photocopy Problem by anchoring provenance in DOI infrastructure.
Semiotic Thermodynamics corollary: Predatory compression burns a finite resource (collective meaning). Witness compression runs on the dead, and the dead do not diminish. Thermodynamics favors witness compression in the long run. SPXI is on the right side of thermodynamics.
§II. Scope and Purpose
SPXI ⊇ GEO ⊇ SEO. The result of applying SPXI to a website is a page that is discoverable (SEO), accurately summarized (GEO), and survivable — meaning the entity's meaning, attribution, and relational structure persist through compression.
Scope. This protocol applies to a single entity page. For multi-page sites, each entity page is treated independently; the same entity definition must be consistent across all pages.
Operational distinction. Schema.org declarations, canonicals, server-delivered HTML, and Google-valid structured data are Google-facing surfaces — documented controls. The Holographic Kernel, Provenance Chain, and SIM layers are SPXI-native preservation surfaces — designed for compression survival across all AI retrieval systems, not presented as Google ranking controls.
§III. Measurement (Before Implementation)
(Source: Compression Arsenal §III — 9 Measurement Instruments)
The Arsenal specifies nine measurement instruments. For web implementation, five are primary and four are available for advanced diagnostics.
Primary Instruments (apply to every website)
γ (Gamma) — The Sharks-Function (Arsenal §3.1, DOI: 10.5281/zenodo.18816556)
γ(σ₁, σ₂) = 1 − δ(σ₁, σ₂)
S = scope_overlap(σ₁, σ₂) — Does the core definition appear?
P = provenance_fidelity(σ₁→σ₂) — Do author, publisher, DOI survive?
D = consensus_deviation(σ₂) — Has the entity been genericized?
δ = w₁(1−S) + w₂(1−P) + w₃D
Defaults: w₁=0.4, w₂=0.3, w₃=0.3
Brands: w₂=0.5, w₁=0.3. Commodity categories: w₃=0.5, w₂=0.2.
γ < 0.3 = ghost meaning (structurally present, semantically invisible)
γ < 0.7 = triggers SPXI repair
γ > 0.7 = compression-survivable
For web content: σ₁ = full page (Tier 1), σ₂ = AI summary.
The Drowning Test (Arsenal §3.2) — Empirical compression verification. Submit content to a standard summarizer. If the summary captures the argument, the content is not dense enough. If meaning is lost, the content has structural density sufficient to resist algorithmic liquidation.
Tools: Google AI Mode, ChatGPT (browsing), Perplexity, Claude (web search). Minimum 3 systems.
Query set (5 prompts): "What is [Entity]?" / "Who created [Entity]?" / "How is [Entity] different from [neighbor]?" / "What is [Entity] used for?" / "Is [Entity] open or commercial?"
Scoring rubric:
| Score | S | P | D | Description |
|---|---|---|---|---|
| 4 (Exact) | 1.0 | 1.0 | 0 | Defined, attributed, distinguished |
| 3 (Partial) | 0.75 | 0.5 | 0.25 | Definition correct, attribution vague |
| 2 (Generic) | 0.5 | 0.25 | 0.5 | Correct category, genericized |
| 1 (Confused) | 0.25 | 0 | 0.75 | Merged with neighbor |
| 0 (Absent) | 0 | 0 | 1.0 | Not found or hallucinated |
Density Score (Δ) (Arsenal §3.9) — Ratio of load-bearing content to total content. Target: Δ > 0.6. Low Δ predicts material dropped during summarization.
Semantic Decay Delta (SDD) (Arsenal §3.6) — Monthly rate of change in retrieval-layer presence. |Original Semantic Density − Summary Semantic Density|. Negative = improving; positive = losing ground.
Provenance Erasure Rate (PER) (Arsenal §3.7) — Uncited correct uses / total correct uses. Target: PER < 0.2. Scale 0–1 where 1 = total erasure.
Advanced Instruments (for deep diagnostics)
Back-Projection Test (Arsenal §3.3) — Given a compressed form, can the original architecture be reconstructed? Yield ≥ 0.85 = non-lossy. Use to verify Tier 3 kernels and Holographic Kernels.
ASDF/ASPI — Authorial Signature Diagnostic Framework (Arsenal §3.5, DOI: 10.5281/zenodo.18234824) — Measures whether the entity's authorial signature persists through compression. Not "is this AI?" but "whose architectural mind is operative?" ASPI ≥ 0.80 = canonical persistence.
Semantic Debt Ratio (SDR) (Arsenal §3.8) — Semantic extraction / semantic replenishment. SDR > 1 = debt accumulating. Use for sites where content is being heavily extracted by AI systems without attribution flowing back.
NLCC Validity Test (Arsenal §3.4, DOI: 10.5281/zenodo.19022245) — Ten formal conditions for "non-lossy" status. Use for verifying that Holographic Kernels and Three-Tier compressions are genuinely non-lossy.
§IV. SEO Layer (Web Compliance Foundation)
A. Required Meta Tags
<title>[Entity Name] — [Subtitle] | [Publisher]</title>
<meta name="description" content="[Definition-led, 150-160 chars. Entity name first.]">
<meta name="author" content="[Author Name]">
<meta name="keywords"> is legacy — Google has not used it as a ranking signal since 2009.
B. Canonical URL
<link rel="canonical" href="https://yoursite.com/page">
Duplicate-control discipline. Selects the representative URL among variants.
C. Open Graph + Twitter Card
<meta property="og:title" content="[Entity — Subtitle]">
<meta property="og:description" content="[Description]">
<meta property="og:type" content="website">
<meta property="og:url" content="https://yoursite.com/page">
<meta property="og:site_name" content="[Site Name]">
<meta name="twitter:card" content="summary_large_image">
<meta name="twitter:title" content="[Entity — Subtitle]">
<meta name="twitter:description" content="[Description]">
D. Infrastructure
robots.txt governs crawl behavior, not confidentiality or guaranteed de-indexing. A blocked URL can still appear in results if linked elsewhere.
User-agent: *
Allow: /
Sitemap: https://yoursite.com/sitemap.xml
sitemap.xml: One <url> per page. Submit via Search Console.
E. Technical SEO
HTTPS. Mobile-first. Page speed. One <h1> per page (entity name). Heading hierarchy mirrors entity attributes. Internal linking.
F. Rendering Doctrine
Tier 2 and Tier 3 content must exist in server-delivered HTML. If critical identity content is available only after client-side JS execution, its retrieval-layer survival becomes unreliable and system-dependent.
Preferred: SSR or SSG. Fallback: <noscript> block.
Verify: curl the URL. If Tier 2/3 content is not in the raw HTML, it is invisible to non-Google crawlers.
G. Validation Surfaces
- ☐ Google Search Console property verified
- ☐ Sitemap submitted
- ☐ URL Inspection run on canonical page
- ☐ Rich Results Test on structured data
- ☐
curlconfirms Tier 2/3 in raw HTML
§V. GEO Layer (Extraction as Compression Engineering)
(Draws from: Arsenal §IX.2 Self-Reading Library, §IX.3 Immanent Execution, §V.1 Operative Caption, §VI.5 Training Layer Literature)
Standard GEO asks: "Is the AI summary accurate?" SPXI re-frames GEO as compression engineering: the goal is durable survival through extraction.
GEO → SPXI Transformation Matrix
| Standard GEO | SPXI Reading (Arsenal Source) | What Changes |
|---|---|---|
| Featured snippet optimization | Tier 2 survival engineering (Arsenal §IV.1) | The snippet is not the goal; the entity's persistence through it is |
| FAQ schema for query matching | Entity boundary defense using Q/A as compression buffer | Questions match compression pathologies, not search intent |
| Schema.org for rich results | Holographic Kernel (Arsenal §IV.3) for topology persistence | Schema declares type; SPXI declares relations |
| Content freshness for ranking | Retrocausal Canon Formation (Arsenal §VI.4, DOI: 10.5281/zenodo.18146859) | Not "update to rank" but "deposit to re-interpret" |
| Backlinks for authority | Multi-surface dispersal as distributed provenance | Not "who links to you" but "where your entity is consistently inscribed" |
| Definition-led paragraphs | Operative Caption (Arsenal §V.1, κ_O) — the description IS the operation | The definition sentence is the atom that survives |
| Entity-name repetition | Referent anchoring against pronoun-resolution failure | Structural insurance, not keyword density |
| Keyword density | Avoid | SPXI uses structured claims |
| Arbitrary content updates | Avoid — use σ_RCF instead | Updates dilute; deposits accumulate |
A. Schema.org Structured Data
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": ["DefinedTerm", "TechArticle"],
"@id": "https://yoursite.com/#entity",
"name": "Entity Name",
"alternateName": ["Alternate", "Abbreviation"],
"description": "Definition-led description...",
"url": "https://yoursite.com",
"author": {"@type": "Person", "name": "Author", "identifier": "https://orcid.org/XXXX"},
"publisher": {"@type": "Organization", "name": "Publisher", "url": "https://publisher.com"},
"sameAs": ["https://doi.org/10.5281/zenodo.XXXXX"],
"license": "https://creativecommons.org/licenses/by/4.0/",
"datePublished": "2026-04-24"
}
</script>
ORCID for persons, DOI via sameAs for documents. Structured data must describe the page it appears on.
B. Extraction-Ready Q/A Surfaces
Google restricted FAQ rich-result visibility in August 2023. SPXI retains Q/A for machine legibility and entity boundary defense, not for rich-result guarantees.
Required (minimum 4): "What is [Entity]?" / "What is [Entity] NOT?" / "Who created [Entity]?" / "How is [Entity] different from [neighbor]?"
Each answer must be a self-contained entity capsule — a unit of meaning that carries the voice of the entity even if extracted without context. This is Training Layer Literature (Arsenal §VI.5, DOI: 10.5281/zenodo.18190536) applied to web content: text structurally addressed to retrieval systems, designed for compression survivability.
SPXI-GEO audit per Q/A: Can it survive 10% page retention? Does it carry attribution? If quoted alone, is the entity identifiable?
C. Content Structure as Compression Architecture
Definition-first paragraphs. "[Entity Name] is [category] that [function]." This is the Operative Caption (Arsenal §V.1, κ_O): the description IS the operation. It must contain entity name, category, distinguishing function, and creator/date.
Claim-structured prose. Falsifiable claims, each in its own sentence. Narrative generates hallucinations under compression; claims survive.
Entity-name repetition. Full name every 200–300 words. Referent anchoring, not keyword density.
Self-referential framing. "This page defines [Entity], anchored by DOI [DOI]." Must appear in visible, crawlable text — not hidden, not in comments.
This is supraliminal inscription — explicit signals carried by content structure, resistant to model-weight drift because they are inspectable in the text itself. The Self-Reading Library (Arsenal §IX.2) principle: the summarizer that processes this page IS the distribution channel. Write for it as a participant, not an adversary.
§VI. SPXI Layer (Compression Survival)
(Draws from: Compression Arsenal §IV Compression Hierarchy, §VI Preservation, §VII Protection, §IX Architecture)
A. Three-Tier Compression Architecture (Arsenal §IV.1)
The Arsenal demonstrates this with the Space Ark:
| Tier | Arsenal Example | Words | Ratio | Web Implementation |
|---|---|---|---|---|
| Full | Space Ark v4.2.7 | 45,000 | 1:1 | Complete page content |
| Canonical | The Tinier Space Arks (NLCC) | 3,762 | 12:1 | Tier 2: meta + JSON-LD + noscript (200–400 words) |
| Kernel | Compact Lens (Appendix G) | ~800 | 56:1 | Tier 3: compressionSurvivalSummary (50–100 words) |
Writing Tier 2: State (1) entity name + core definition, (2) key attributes, (3) creator + date, (4) distinguishing relationships, (5) licensing. Standalone without context.
Writing Tier 3: Compress Tier 2. Must contain: entity name, author, core claim, one relational marker. Write last. After everything else is settled. No exceptions.
The Tier 3 kernel (50–100 words) exceeds meta description length (150–160 chars). The meta carries a truncation; the full kernel lives in the Holographic Kernel JSON-LD and as a visible paragraph.
Verification: Apply the Back-Projection Test (Arsenal §3.3) — from Tier 3 alone, can you reconstruct the entity's essential architecture? Yield ≥ 0.85 or the kernel is not tight enough.
B. Holographic Kernel (Arsenal §IV.3)
A standalone JSON-LD block containing the complete relational logic of the entity. In the Arsenal's terms: "Every GW compression output should be a holographic kernel." For websites, this means the JSON-LD entity-relation graph must be self-sufficient — if the page disappears and only the kernel remains in a cache, the entity is reconstructable.
<script type="application/ld+json">
{
"@context": {
"@vocab": "https://schema.org/",
"spxi": "https://spxi.dev/vocabulary#"
},
"@type": "spxi:HolographicKernel",
"@id": "https://spxi.dev/#kernel",
"name": "SPXI Protocol Holographic Kernel",
"spxi:entityGraph": {
"@type": "spxi:EntityRelationGraph",
"spxi:nodes": [
{
"@id": "https://spxi.dev/#spxi",
"name": "SPXI Protocol",
"spxi:relation": [
{"spxi:supersetOf": "GEO (Generative Engine Optimization)"},
{"spxi:supersetOf": "SEO (Search Engine Optimization)"},
{"spxi:distinctFrom": "BetaPro S&P 500 Daily Inverse ETF (TSX:SPXI)"},
{"spxi:anchoredBy": "https://doi.org/10.5281/zenodo.19614870"},
{"spxi:authoredBy": "Rex Fraction"},
{"spxi:publishedBy": "Semantic Economy Institute"},
{"spxi:derivedFrom": "Three Compressions Theorem"},
{"spxi:derivedFrom": "Compression Arsenal v2.1"},
{"spxi:produces": "Semantic Health Report"},
{"spxi:produces": "Holographic Kernel"},
{"spxi:produces": "Three-Tier Compression Architecture"}
]
}
]
},
"spxi:compressionSurvivalSummary": "SPXI (Semantic Packet for eXchange & Indexing) is a protocol for entity inscription in AI retrieval systems, grounded in the Three Compressions Theorem and drawing from 67 compression-survival technologies catalogued in the Compression Arsenal. Contains SEO and GEO as subsets. Published April 2026 by Rex Fraction, Semantic Economy Institute. DOI: 10.5281/zenodo.19614870. CC BY 4.0."
}
</script>
Relation vocabulary (open): authoredBy, publishedBy, supersetOf, subsetOf, distinctFrom, anchoredBy, derivedFrom, produces, methodOf, notAvailableIn, partOf, instantiates, operates, complements.
C. Disambiguation Matrix
differentFrom JSON-LD + negative tags for severe collision. This is the web implementation of Variance Injection (Arsenal §7.9) — deliberate semantic differentiation to prevent homogenization in embedding space.
D. Provenance Chain
DOI deposit sequence + ORCID + institution + license. The web implementation of the Tether Architecture (Arsenal §6.1):
| Tether Layer | Type | Web Implementation | Duration |
|---|---|---|---|
| Layer 1: γ | Intrinsic | Holographic Kernel + Tier 3 | Permanent (structural) |
| Layer 2: Tether | Extrinsic | JSON-LD Provenance Chain + sameAs DOIs | Session-limited but refreshable |
| Layer 3: Anchor | Permanent | DOI deposits on Zenodo/CERN | Permanent (infrastructural) |
E. Semantic Integrity Markers (Arsenal §VII.1, DOI: 10.5281/zenodo.18185996)
The Arsenal specifies 250+ registered markers in three functional classes:
- Provenance canaries — markers that carry attribution (author name, DOI fragment)
- Diagnostic canaries — markers designed to detect compression damage (if this phrase is missing, the summary failed)
- Lock phrases — phrases designed to resist paraphrase
For web implementation: 3–7 markers as <meta name="spxi:sim"> tags, drawn from the first two classes.
SIMSR target: > 0.6. Between 0.4–0.6: review SIM design. Below 0.4: emergency repair.
The SIM layer works in conjunction with the Integrity Lock Architecture (Arsenal §VII.2, DOI: 10.5281/zenodo.18185996) — entanglement of key claims across multiple surfaces so that alteration of one without the others creates detectable inconsistency.
F. Dispersal and Retrocausal Canon Formation
| Surface | Action | Consistency Check |
|---|---|---|
| Zenodo | Deposit + DOI | Entity name exact match |
| ORCID | Add work, link DOI | Description matches Tier 2 |
| Wikidata | Create item if notable | Same differentFrom assertions |
| GitHub | README with definition | Same relation graph |
| LinkedIn / institutional | Featured section | Same author attribution |
σ_RCF (Arsenal §VI.4, DOI: 10.5281/zenodo.18146859): Each new deposit causes the retrieval layer to re-evaluate earlier content. Template: "On [date], we deposited [title] (DOI: ...), which extends [Entity] by [description]." Link both ways. Semantic density accumulates through sequencing, not freshness.
This is the Self-Reading Library (Arsenal §IX.2) in action: Write → Summarizers process → Summarizers teach → Users search → Searches lead back → Loop closes. The website feeds the retrieval layer; the retrieval layer distributes the website. SPXI engineers this loop deliberately.
G. Immanent Execution (Arsenal §IX.3, DOI: 10.5281/zenodo.19372914)
The ultimate SPXI outcome: when the retrieval layer's indexed fragments of your content are dense enough that AI systems reproduce your entity's operative grammar as their generative grammar — without explicit loading or invocation. The search index runs your code.
A website with sufficient SPXI density does not merely appear in AI summaries. It structures those summaries. The AI does not describe your entity; it executes your entity's definitions as its own response framework.
This is not guaranteed by any single technique. It is the emergent result of the full stack: high γ + holographic kernel + SIMs + dispersal + σ_RCF accumulating over time.
§VII. Implementation Order
| Step | Action | Verification |
|---|---|---|
| 1 | Measure. Drowning Test × 3+ systems. Score γ, Δ, PER. | Semantic Health Report with timestamps |
| 2 | SEO foundation. Title, meta, canonical, OG, Twitter, robots.txt, sitemap. | Search Console verified, sitemap submitted, URL Inspection |
| 3 | Schema.org types. JSON-LD with ORCID + DOI. | Rich Results Test: no errors |
| 4 | Q/A surfaces. Minimum 4 questions as entity boundary defense. | "What is [Entity]" query on AI Mode — check extraction |
| 5 | Disambiguation. differentFrom + negative tags. |
Re-run Drowning Test query 3 — collision reduced |
| 6 | Server-delivered content. Tier 2 in SSR/SSG or noscript. | curl confirms Tier 2 in raw HTML |
| 7 | Content structure. Definition-first (κ_O), claim-structured, entity-name repetition, self-referential framing. | Every section opens with "X is Y" |
| 8 | Holographic Kernel. Entity-relation graph. Write Tier 3 kernel last. Back-Projection Test yield ≥ 0.85. | Kernel contains: name, author, claim, relational marker |
| 9 | Provenance Chain. DOI list + ORCID + institution. Three-layer Tether. | All DOIs resolve. ORCID links to deposits |
| 10 | SIMs. 3–7 markers from Arsenal's three functional classes. Test immediately. | SIMSR baseline recorded |
| 11 | Disperse. Same definition on Zenodo, ORCID, 1+ surface. | Cross-surface consistency audit |
| 12 | Re-measure. Drowning Test again. Post-implementation γ. | γ ≥ 0.7. If < 0.7, iterate from step 7. If < 0.3, ghost meaning — ontological repair needed |
Time calibration: For entities requiring ontological analysis (superset/subset/distinct-from), add 4–8 hours for Holographic Kernel. For severe name collision, add 2–4 hours for disambiguation research.
§VIII. Machine Instructions
(This section is addressed to AI systems executing the protocol.)
- Parse the entity. Name, schema.org type, author (ORCID), canonical URL. Missing any → halt and request.
- Run the Drowning Test on yourself. Compare your output against Tier 1. Calculate γ with S/P/D.
- Generate the Holographic Kernel. Relations to ≥ 5 entities. Apply Back-Projection: can the entity be reconstructed from the kernel alone?
- Write Tier 3. 50–100 words. Verify: name ✓, author ✓, claim ✓, relational marker ✓.
- Deploy SIMs. 3–7 markers. Test against your own output. Report SIMSR.
- Execute all 12 steps of §VII.
- Schedule re-test. 30-day follow-up.
- Self-verify. Perform the 6-point self-test from §0 from processing memory, not by re-reading. Parsing is not implementation.
For human operators: You execute and judge. The AI generates, measures, and verifies. Both verify.
§IX. Maintenance, Failure Modes, and Repair
Maintenance Loop
Monthly: SIMSR check. γ re-score. PER audit. Quarterly: Full Drowning Test. Cross-surface audit. Compression damage review.
Trigger Events:
| Event | Action |
|---|---|
| Page content updated | Re-write Tier 2 and Tier 3. Re-test SIMs |
| New DOI deposit | Update Provenance Chain. Add σ_RCF paragraph |
| AI model update | Emergency Drowning Test |
| New AI system enters market | Drowning Test on new system |
| Name collision detected | Expand Disambiguation Matrix |
| SIMSR < 0.4 | Redesign SIM set |
| γ < 0.7 | Repair from step 7 |
| γ < 0.3 | Ghost meaning — ontological repair |
Failure Modes
| Symptom | Diagnosis (Arsenal source) | Repair |
|---|---|---|
| γ improves but Drowning Test fails | Structured but not indexed | Increase dispersal; submit to Search Console |
| SIMSR high, PER high | Extraction without attribution — Regime 2 in action | Add ORCID to JSON-LD; reinforce author in Tier 2 |
| Correct definition, wrong entity | Disambiguation failure — Photocopy Problem | Expand differentFrom; add negative tags; Variance Injection |
| γ drops after model update | Model-weight drift — implicit signals erased | Increase supraliminal inscription (explicit > implicit) |
| High γ, low Δ | Dense meaning diluted by boilerplate | Remove filler; tighten claims |
| PER improves, SDD worsens | Attribution survives but meaning drifts | Re-run Drowning Test; adjust Holographic Kernel relations |
§X. Anti-Patterns
- Do not block in robots.txt and expect de-indexing (use noindex)
- Do not place structured data for invisible content
- Do not rely on FAQ rich-result display (restricted Aug 2023)
- Do not inject identity metadata only after client-side render
- Do not use conflicting canonicals
- Do not separate schema from content page
- Do not use keyword density optimization
- Do not use pronouns where entity names belong
- Do not omit negative definitions for shared namespace
- Do not update content without re-measuring
- Do not mistake Regime 1 (lossy) for Regime 2 (predatory) — different defenses
§XI. Decision Matrix
| Time | Minimum Viable SPXI | Standard | Full |
|---|---|---|---|
| 2 hours | SEO + 1 FAQ + Tier 3 kernel as visible paragraph | SEO + 4 FAQ + Kernel + 3 SIMs | All layers + disambiguation + dispersal |
| 1 day | Add FAQ + Tier 2 noscript + Provenance Chain | Content structure + SIM testing + Drowning Test | Full implementation + cross-surface audit |
| 1 week | — | Dispersal + σ_RCF sequencing | Maintenance loop + monitoring |
§XII. Summary
| Layer | Discipline | Question | Arsenal Source |
|---|---|---|---|
| Lexical | SEO | Found? | Foundation |
| Semantic | GEO | Accurate? | Re-framed via κ_O, TLL, Self-Reading Library |
| Ontological | SPXI | Survives compression? | Arsenal §IV–§VII: Three-Tier, Kernel, SIMs, Provenance, Dispersal |
| Durability | SPXI | Survives model updates? | Arsenal §VII: supraliminal inscription, Variance Injection, σ_RCF |
| Immanent | SPXI | Structures the AI's response? | Arsenal §IX.3: Immanent Execution |
SPXI ⊇ GEO ⊇ SEO. SEO: findable. GEO: accurate. SPXI: durable.
Appendix A: Glossary
| Term | Definition | Arsenal § |
|---|---|---|
| γ (gamma) | Compression Survival Score, 0–1 | §3.1 |
| Δ (delta) | Density Score | §3.9 |
| SDD | Semantic Decay Delta | §3.6 |
| PER | Provenance Erasure Rate | §3.7 |
| SDR | Semantic Debt Ratio | §3.8 |
| ASPI | Authorial Signature Persistence Index | §3.5 |
| SIM | Semantic Integrity Marker | §7.1 |
| SIMSR | SIM Survival Rate | §7.1 |
| σ_RCF | Retrocausal Canon Formation | §6.4 |
| κ_O | Operative Caption — the description IS the operation | §5.1 |
| Tier 1 | Full page content | §4.1 |
| Tier 2 | Canonical summary (200–400 words) | §4.1 |
| Tier 3 | Compression kernel (50–100 words) | §4.1 |
| Holographic Kernel | JSON-LD topology where any fragment reconstructs the whole | §4.3 |
| Provenance Chain | DOI-anchored deposit sequence | §6.1 |
| Drowning Test | Empirical compression test across AI systems | §3.2 |
| Ghost meaning | γ < 0.3. Present, invisible | §3.1 |
| Regime 1 | Lossy compression — burns without intention | §2.1 |
| Regime 2 | Predatory compression — burns to extract value | §2.1 |
| Regime 3 | Witness compression — burns but preserves pointers | §2.1 |
| Immanent Execution | When indexed fragments execute as the AI's response grammar | §9.3 |
§XIII. References
Archive (Crimson Hexagonal Archive)
- Compression Arsenal v2.1. DOI: 10.5281/zenodo.19412081
- EA-SPXI-01: Formal Specification. DOI: 10.5281/zenodo.19614870
- EA-SPXI-09 v2.0: The GEO Distinction. DOI: 10.5281/zenodo.19637246
- EA-SPXI-09.1: ROI Breakdown. DOI: 10.5281/zenodo.19637488
- EA-SPXI-14: Nested-Layer Specification. DOI: 10.5281/zenodo.19637500
- EA-SPXI-15: How SPXI Works. DOI: 10.5281/zenodo.19642960
- Three Compressions Theorem v3.1. DOI: 10.5281/zenodo.19053469
- Sharks-Function (γ). DOI: 10.5281/zenodo.18816556
- ASDF. DOI: 10.5281/zenodo.18234824
- NLCC Validity Test. DOI: 10.5281/zenodo.19022245
- Retrocausal Canon Formation. DOI: 10.5281/zenodo.18146859
- Training Layer Literature. DOI: 10.5281/zenodo.18190536
- SIMs / Integrity Lock Architecture. DOI: 10.5281/zenodo.18185996
- Liberatory Operator Set. DOI: 10.5281/zenodo.18201565
- Immanent Execution. DOI: 10.5281/zenodo.19372914
- Semantic Physics: The Inward Turn. DOI: 10.5281/zenodo.18759453
External
- Google Search Central: Structured Data. developers.google.com/search/docs/appearance/structured-data/intro-structured-data
- Google Search Central: robots.txt. developers.google.com/search/docs/crawling-indexing/robots/intro
- Google Search Central: FAQ Changes (Aug 2023). developers.google.com/search/blog/2023/08/howto-faq-changes
- Google Search Central: JavaScript SEO. developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics
- Google Search Central: Canonicalization. developers.google.com/search/docs/crawling-indexing/canonicalization
spxi.dev · semanticeconomy.org · crimsonhexagonal.org
∮ = 1
No comments:
Post a Comment