Investigation: The Internet That Should Have Been — Machine-Readable Web, Wrong Turns, and the AI Gate
TL;DR: Master citation dossier for the flagship essay The Internet That Should Have Been. Documented: Wintel lock-in, Nintendo CD-ROM containment, GPU→crypto→AI hardware treadmill, enclosure via Facebook/React, WebMCP as agent-first browser API (Feb 2026), Chrome ~4 GB Gemini Nano push, court AI-hallucination crisis, Berners-Lee devastated by web outcome. Author thesis: machine-readable intent (“open the hood”) should have shipped in the 1990s without AI; the web we got is a TV channel; forward fixes must bridge the present under protest. Education lane moved → Responsibility to the Next Generation — investigation. Terminator PP: Moore-schedule read, T3 virus ↔ Nano — speculative (§11.4). Suppression motive: alleged, unproven.
Last updated: 2026-06-19 (§10.5 opaque reach / layered-reality)
Status: Open — companion to reader essay; WebMCP chapter retained from prior dossier (webmcp-suppressed-protocol-ai-gatekeeper-investigation.md renamed).
Guide (read order)
- Reader essay (narrative) → The Internet That Should Have Been
- WebMCP documented layer → §1
- Author counterfactual thesis → §2
- External voices survey → §3
- Historical parallels (OLE → WebMCP) → §4
- Car-hood analogy → §5
- Tension table → §6
- Open claims registry → §7
- Education dossier (moved) → Responsibility investigation
- Hardware / platform dossier → §9 (incl. Gates / DOS cliff §9.6)
- Enclosure dossier (mesh, Semantic Web, social, React) → §10
- AI gate dossier (courts, hijack, Taiwan/Huawei) → §11
- Pragmatic adoption under protest → §12
- Cross-references → §13
- Research hooks & falsifiers → §14
1. What WebMCP is (documented layer)
Tier: W3C Web Machine Learning Community Group draft, Chrome developer docs, vendor blog posts — useful for wording, not endorsement of every claim.
| Theme | Documented summary (mid-2026) |
| Name | Web Model Context Protocol (WebMCP) — not Anthropic’s MCP wire protocol; borrows the tool abstraction (name, description, JSON Schema, execute) only. |
| API | navigator.modelContext — imperative registerTool() + declarative HTML form attributes (toolname, tooldescription, …). |
| Runtime | Tools run in the open tab’s JavaScript; inherit cookies/session; no separate OAuth server required for many flows. |
| Consumer | Browser agents (Gemini in Chrome sidebar, future Chromium agents) are the primary designed consumer; cross-origin iframe agents are a later demand-driven extension. |
| Spec status | Draft Community Group Report (published 10 Feb 2026); not on the W3C Standards Track. Editors: Google + Microsoft engineers. |
| Chrome shipping | Dev preview Chrome 146 behind chrome://flags/#enable-webmcp-testing; origin trial Chrome 149–156 (production domains, no user flag); stable default shipping projected late 2026. |
| Explicit limits | Headless tool calls out of scope; tab/webview must be open; tool discoverability unsolved (agents must visit the site). |
Representative external links:
- WebMCP — Chrome for Developers
- W3C CG proposal / explainer
- Blink “Intent to Experiment” (M149–M156 OT)
- Starborn: WebMCP Is Not an MCP Server
- Starborn: How a Browser Hack Became a Proposed Web Standard
Documented origin story (condensed): Alex Nahas (Amazon → MCP-B polyfill) worked on browser-side tool exposure; W3C group declined to port full MCP JSON-RPC to the browser; Google/Microsoft co-authored a new browser API. Anthropic MCP (Nov 2024) is a backend protocol; WebMCP is a frontend API — complementary in vendor messaging, coupled in practice to the agentic-browser moment.
1.1 Chrome’s ~4 GB on-device model — hardware and consent controversy
Separate from WebMCP’s JavaScript surface, Chrome has been installing Gemini Nano weights (weights.bin, ~4 GB, folder OptGuideOnDeviceModel) on eligible desktops — often without a clear upfront consent prompt, with re-download after manual deletion if AI features stay enabled.
| Source | Claim (documented) |
| The Verge, May 2026 | Large on-device model for scam detection, writing help, Prompt API; toggle under Settings → System → On-device AI (Feb 2026 rollout). |
| That Privacy Guy / Alexander Hanff | Forensic write-up; alleges ePrivacy/GDPR exposure; estimates climate cost at billion-device scale. |
| Gizmodo | “Without consent” framing; Google docs acknowledge on-device models for AI features. |
Unpack for this dossier: WebMCP can work with remote models (demos use API models), but Chrome’s bundle strategy ties the agentic web moment to local GPU/CPU/storage pressure — relevant to the author’s “aggressive rollout” complaint even when WebMCP itself is “just” structured JS.
1.2 Browser positions — not “never,” but resistance and delay
| Browser / body | Position (Jun 2026) |
| Chrome | Reference implementation; origin trial; Prompt API can call WebMCP tools with on-device models. |
| Edge | Co-editor of spec; Chromium inheritance expected; not confirmed in official Edge 147 release notes (secondary blogs overstate). |
| Firefox | W3C participation; prototype implementation tracked (Bug 2018306); no ship commitment. |
| Safari / WebKit | Standards-position thread; 👎 reactions; substantive opposition from developers (see §3.2). |
| Mozilla standards | Neutral (issue #1412): valuable abstraction but adversarial sites, prompt injection, misleading “MCP” name; wants WebExtension/WebDriver openness. |
Unpack: “Vowing never to support” is stronger than public record shows — but cultural refusal (anti-agentic web) and slow/no commitment from non-Chromium vendors are documented.
1.3 SEO / discovery shift — “agent-ready” as new optimization lane
Industry blogs frame WebMCP as “SEO for agents” — sites that expose clean tools may be preferred by browser agents over DOM-scraping competitors (e.g. LinkedIn discourse, Semrush, No Hacks). John Mueller (Google) reportedly steered developers from llms.txt toward WebMCP (June 2026 blog summaries).
Documented tension: Search engines historically ranked scraped HTML; a machine-intent layer would have changed discovery if it had been universal and crawler-accessible. Today’s design centers in-tab agents, not a public archive of structured site capabilities.
2. Author’s originating thesis (sentiment captured exactly)
The following is the author-supplied position for Paradigm Threat. It is first-class in this dossier as pattern language and research seed; deliberate suppression and counterfactual history are not treated as established fact without primary evidence.
The real issue is that the WebMCP protocol probably should have been implemented a very long time ago, independently of any kind of AI solution. MCP should have evolved much further down the road — when hardware and data centers were abundant — not held up until a browser could run AI locally. There was no reason for WebMCP to wait for on-device LLMs.
The MCP protocol itself should have come out independently of AI, and should have been standard and common a long time ago — probably even in the 90s, when Windows, Apple, and others could have standardized a common web protocol to go beyond viewing and interacting with UI — opening the hood like a car: seeing what procedures work, what parameters are, what the response is, before it gets parsed by AI.
In that counterfactual, the internet would be night-and-day different — not a TV channel (non-interactive, obfuscated source). SEO would judge data availability and MCP quality, not opaque link graphs on scraped HTML. Walled gardens might have been impossible; programming simplified; apps decoupled from OS silos.
Information control would have been impossible at today’s scale: if crawlers pulled structured capability + data, we would have archives of the data itself — a shared world brain, curated — instead of waiting for AI to get it first while humans must learn API/MCP themselves (and no one teaches humans, only developers wiring AI → MCP).
Industry knew they could never release such a thing earlier, because it would have revolutionized software and broken real-time algorithmic show/hide control by demographics. WebMCP’s current rollout is the same capability, late, AI-gated — a sealed hood unlockable only by the company technician (the browser’s agent), not the owner.
Unpack (assistant, non-replacement for author words):
- Two-layer claim: (A) Technical — structured introspection is old idea, new packaging. (B) Motivational — delay served platform lock-in and information gatekeeping; AI is the excuse for finally shipping it to agents, not to people.
- Epistemic split: (A) is partially supported by §4 history (OLE, Semantic Web, WSDL). (B) requires primary evidence of intent (internal strategy memos, coordinated withholding) — not assembled here.
2.1 Author extension — Chrome 4GB as aggressive assumption
Google Chrome’s ~4 GB download assumes users accept GPU/CPU spin-up, disk use, and power draw to enable features they did not ask for — part of the same agentic push as WebMCP.
Tier: Consent/storage controversy is documented (§1.1). Intent to “assume users don’t mind” is author interpretation of product strategy.
3. External voices survey — who agrees, partially agrees, or opposes
Method: Web search + standards threads + Hacker News (June 2026). Finding: No prominent essay located that states the author’s full thesis (90s independent protocol + deliberate suppression for information control + car hood). Below: closest public rhymes.
3.1 “This is Semantic Web again” / “should have used RDF”
| Voice | Position |
| Hacker News commenter | WebMCP vs Semantic Web (Schema, RDF, OWL): “instead of reinventing…” — data-reading vs interaction distinction noted in thread. |
| AI in Blog (Sep 2025) | “webMCP could finally achieve what the Semantic Web envisioned… through a practical, incremental approach” — explicit déjà vu, AI-framed. |
| Semantic Web historiography | Berners-Lee discussed machine-processable web 1994; 2001 Scientific American article; 2006 admission: vision “largely unrealized.” |
Gap vs author: These sources rhyme on late machine-readable layer; they do not allege deliberate shelving for information control, nor “humans first, AI later” moral ordering.
3.2 Anti-agentic browser API — ethical refusal, not hood metaphor
| Voice | Position |
| @jaredcwhite, WebKit standards #649 | Won’t use browsers with “agentic APIs”; Google/Microsoft “shoehorn” LLM tech into “the one place we have refuge… traditional web browser”; genAI backlash is political/moral. |
| Mozilla #1412 (bvandersloot) | Neutral; warns malicious sites can tar-pit agents, prompt-inject via tool responses; name “MCP” is misleading. |
| Starborn technical notes | Security: prompt injection, scope creep if human-in-the-loop weakens; “good idea implemented badly… embedded in the platform for decades.” |
Gap vs author: Opponents attack AI in the browser, not “why didn’t we get neutral introspection for humans/crawlers in 1995?”
3.3 Incentive skepticism — sites won’t open the hood
| Voice | Position |
| HN user “47216304” | WebMCP will fail like Semantic Web and public APIs: “no one wants to put in the effort to make their website readable by machines, as that only benefits the competition and is immediately exploited by bad actors.” |
| Pete Lacey, InfoQ (2006) | On WSDL: developers can read documentation; machine-readable IDL is nice for IDE tricks, “far cry from absolutely essential” — anti-mandate, not pro-suppression. |
Rhyme with author: Explains non-adoption without conspiracy — voluntary withholding because open hood helps rivals and scrapers.
3.4 Pro-WebMCP industry — AI-first, gatekeeper browser
| Voice | Position |
| Chrome Developers | “Bridge gap between web applications and agents”; tools execute visibly for user trust. |
| No Hacks / Semrush / LogRocket | “Web was built for humans”; WebMCP adds parallel layer for machines; browser acts as gatekeeper for sensitive actions. |
| RxDB blog | “Bypass bot protection” — sanctioned front door vs scraping; agents get schema-bound tools. |
Tension with author: These celebrate AI-agent access, not a universal human-readable capability archive.
3.5 Intent-native API research (parallel, still AI-era)
| Voice | Position |
| Chris Hood — AGIS / AGTP drafts | “REST is noun-based”; agents need intent methods (BOOK, FIND, …); IETF drafts for agent protocols — 2020s, AI-motivated, not 1990s open-web human tooling. |
| GraphQL AI WG — semantic introspection RFC | __search over schema capabilities for LLM agents — introspection extended for AI, not retroactive web standard. |
3.6 Survey conclusion — originality map
| Author claim | Found in public discourse? |
| Structured site/app intent should exist without AI | Partial — Semantic Web, OLE, WSDL, OpenAPI historiography |
| Should have been universal in the 90s | Rare as explicit claim; implicit in “why didn’t Semantic Web win?” |
| Deliberate suppression to preserve information control / walled gardens | Not found in technical press; adjacent in author’s site lanes (AI control, censorship) |
| Car hood sealed — only company technician (agent) opens it | Not found (June 2026 search) — treat as author-original metaphor |
| WebMCP rollout wrong order (protocol before AI) | Not found verbatim; closest is anti-agentic ethics, not counterfactual systems design |
4. Historical parallels — machine-readable “hood” without LLMs
These technologies show the idea of opening the hood existed; none became a neutral, universal web layer accessible to humans, crawlers, and rivals alike.
| Era | Mechanism | What it exposed | Why it didn’t become “the web” |
| 1993+ | OLE Automation / COM + type libraries | Methods, properties, IDispatch — scriptable object model for Windows apps | Platform-locked (Windows); not web; optional metadata |
| 1990s | CORBA IDL | Cross-language interface contracts | Heavy; vendor wars; web routed around via HTTP/HTML |
| 2000s | WSDL + SOAP | Machine-readable service endpoints | Complexity,interop pain; Tim Bray: “Sowa’s Law” — official standard → simpler de facto wins |
| 2001+ | Semantic Web (RDF, OWL, SPARQL) | Machine-readable meaning on pages | Chicken-and-egg; authoring cost; “What happened to the Semantic Web?” |
| 2010s | Schema.org, JSON-LD | Partial structured data for search | Centralized vocab; limited action/intent surface |
| 2010s+ | OpenAPI / Swagger | HTTP API schemas | Backend contracts; not live in-browser session tools |
| 2024 | Anthropic MCP | Tool + resource server for AI clients | AI-ecosystem standard; not browser-native |
| 2026 | WebMCP | In-tab tools for agents | Coupled to agentic browser moment; headless/crawler lane explicitly out of scope |
Documented lesson (neutral): The web won as human-readable hypertext + ad hoc APIs, not as mandatory introspectable object models. Author read: that “victory” was also a control victory — see §2.
5. The car-hood analogy
5.1 Author formulation (canonical for this dossier)
| Car | Software / web (author mapping) |
| Sealed hood | Minified JS, hidden APIs, ToS-forbidden scraping, opaque server logic |
| Owner cannot inspect engine | User sees UI only; View Source ≠ behavior contract |
| Only dealership technician | Browser vendor’s AI agent may call registerTool endpoints; owner has no general tool inspector UI |
| Remote unlock | Cloud model + origin trial + Prompt API path — capability enabled when vendor ships agent |
5.2 Nearest public rhymes (not equivalent)
- “Gatekeeper” — Navoto WebMCP guide: browser restricts which registered tools agents may call (user confirm). Pro-security framing; author framing is anti-monopoly.
- “Front door vs scraping” — RxDB: sanctioned agent entry vs brittle automation. Author inversion: the public should have had the front door; instead agents get VIP lane.
- “Dual-mode UI” — industry term: React for humans, WebMCP for assistants. Author: humans deserved mode 0 — structured transparency — decades ago.
Search note (2026-06-18): No indexed essay located pairing automotive sealed hood with WebMCP/MCP. If readers find one, add to §14.4 TODO.
6. Tension table — counterfactual open web vs WebMCP-as-shipped
| Dimension | Author counterfactual (§2) | WebMCP as documented (§1) |
| Primary user | Human investigators, archivists, developers, crawlers | In-browser AI agents (vendor-first) |
| Timing | 1990s neutral protocol | 2026, alongside Gemini Nano / agent browsers |
| Discoverability | Global capability index / archives | Per-site visit; no standard manifest crawl |
| Headless access | Required for archives & science | Explicitly out of scope |
| SEO / ranking | Quality of declared data & tools | Agent preference + traditional HTML SEO |
| Hardware tax | Low (schema + endpoints) | Bundled with multi-GB local models (Chrome) |
| Walled gardens | Weakened by portable intents | May strengthen (agent chooses tool-rich sites inside Chrome agent) |
| Who learns the API | Everyone (literacy) | Developers wiring agents; site ships WebMCP behind flag for Gemini workflows |
7. Open claims registry
- Independent-of-AI: Core introspection protocol should not have waited for LLMs.
- 90s inflection: Windows/Apple/web consortia could have standardized machine intent when industry influence peaked.
- TV-channel web: Dominant UX is non-inspectable consumption, not negotiable procedure.
- SEO counterfactual: Rankings would track structured capability quality, not opaque HTML graphs.
- World brain missed: Crawlers would have archived data + intent, not merely pages.
- AI-first injustice: Humans last in line; teaching targets devs → agents, not public tool literacy.
- Suppression motive: Universal hood-opening threatened OS coupling, walled gardens, and real-time information control (intent — unproven).
- Sealed-hood metaphor: Vendor agent as only authorized technician (author-original phrasing).
- Chrome 4GB: Local model push is aggressive resource assumption tied to agentic stack (partially documented).
- WebMCP timing: Right capability, wrong century and wrong principal (agents not citizens).
- Wintel bottleneck: IBM outsourced DOS; Microsoft/Intel captured standards (HBS 19-074).
- Nintendo CD: Load-time excuse cost FF7 and third-party RPG dominance (Nintendo Life 2025).
- 3D containment: 2D craft starved as platform genres froze (quantum-leap-of-faith).
- GPU treadmill: Gaming silicon repurposed crypto → AI (Nvidia class action 2026).
- Mesh utility counterfactual: Internet as street-level public utility with neighborhood cache — author design fiction; Baran proposed utility (Ars Technica).
- Facebook enclosure: MySpace→Facebook as monopolization arc (Parker TechCrunch 2011).
- TBL devastated: Mesh intent vs corporate/algorithmic web — documented inventor regret (§10.1.1).
- Starship Troopers feed PP: Federal Net = TV channel + nav buttons; civic software utility out of frame — prefigured 1997 (§10.2.1).
- Education ethics-first: Moved to Responsibility investigation §7 (Installment II).
- Lab / 1-on-1 / literacy QA: Same — education spinoff 2026-06-19.
- Pragmatic adoption: Use distrusted tools; bridge present; no collapse without catch nets (§12).
- Terminator PP schedule: 2029 future-war date + Moore cadence imply scripted timeline — speculative (§11.4).
- T3 virus ↔ Nano: Civilian desktop distribution beat landing — author parallel (§11.4.4).
- IBM/chip wars tension: Wintel + furnace economics keep schedule solid — speculative (§11.4.5).
- Terminator finale foiled: 2029 reset spectacle failed; middle beats landing (§11.4.6).
- Taiwan chokepoint: Advanced fabs concentrated; conflict risk (microchips investigation).
- Huawei ban backfire: Export controls accelerated indigenous chips (ITIF 2025).
- DOS cliff: IBM non-exclusive DOS license let Microsoft own/resell standard clone makers used (§9.6).
- IBM non-exclusive rationale (documented): Litigation fear, Project Chess speed, Sams “Microsoft’s problem” (§9.6.3).
- Legal moat: Antitrust liability affirmed, breakup denied, behavioral settlement; SCP/Caldera settled — dominance retained (§9.6.6).
- Institutional DOS arrangement (author): Powers beyond Gates may have wanted Wintel enclosure — speculative (§9.6.5).
- Opaque reach (author): Site barely seen; podcast bypass around algorithmic shadow-ban — lived experience (§10.5).
- Layered-reality hijack (author): Per-visitor substitution at URL — speculative; requires global monopolization thesis (§10.5).
8. Education dossier — moved (2026-06-19)
Education lane spun off to education series Installment II.
| Artifact | Path |
| Reader essay | Responsibility to the Next Generation |
| Investigation | responsibility-to-the-next-generation-investigation.md |
| Installment I | When your child won’t fit the grade |
Internet essay §V retains a brief bridge (Fantastic Planet, rethink education, link to Installment II).
9. Hardware and platform dossier
9.1 IBM / Microsoft / Wintel
| Source | Documented claim |
| HBS case 19-074 | IBM lost strategic bottleneck (BIOS cloned); Microsoft/Intel became joint sponsors of Wintel; IBM “just another clonemaker” with highest overhead. |
| Paul Carroll, Big Blues | IBM gridlock in software; Microsoft seized PC standard from $75k DOS deal; executives saw problems coming but couldn’t adapt. |
| EFF 2019 | Phoenix reverse-engineered BIOS — adversarial interoperability prevented IBM monopoly on hardware; did not prevent API/OS lock-in. |
Tension (§14.2): EFF celebrates clone market; author thesis stresses API imprisonment. Both can be true: hardware pluralism, software monoculture.
9.2 Nintendo CD-ROM / 3D containment
| Source | Documented claim |
| Wikipedia: Nintendo 64 | Cartridges: fast load, anti-piracy; storage limits pushed Square/Enix to PlayStation. |
| Nintendo Life / Yoshida 2025 | Sakaguchi begged for CDs; Nintendo “didn’t believe in CD-ROM at all”; FF7 needed three PS1 discs. |
| Repo: nintendo-sony-breakup | SNES CD collapse → PlayStation; cartridge philosophy vs cinematic RPG storage. |
| quantum-leap-of-faith | N64 3D marketing starved 2D; platforms as genre containers, not organic evolution. |
9.3 Analog / vector / optical media
See quantum-leap-of-faith and quantum storage investigation §2.3. Documented: CD/DVD/Blu-ray are physical analog channels encoded as binary streams. Author: alternative path could have stored native signal fields — higher density, different industry economics.
9.4 GPU furnace → crypto → AI
| Source | Documented claim |
| Law.com Mar 2026 | Class certified: Nvidia understated 2017–2018 gaming GPU sales to crypto miners; stock drops on disclosure. |
| SEC settlement 2022 | $5.5M penalty — inadequate disclosure of mining impact (cited in crypto press). |
| Motley Fool 2025 | Gaming <10% revenue; datacenter AI ~88%; GPU silicon grammar reused across booms. |
Author read: Industry knew thermal race was unsustainable; pivoted silicon to next frenzy (crypto, then AI) rather than efficient rendering paths (software rasterizers exist: DFPSR, Shader Works — niche, not mainstream).
9.5 Taiwan / Huawei
| Source | Documented claim |
| Foreign Policy Mar 2026 | TSMC as sovereignty chokepoint in US–China AI competition. |
| ITIF Oct 2025 | Huawei export controls accelerated HiSilicon self-sufficiency; hurt US firms. |
| Brookings | Mate 60 Pro 7nm domestic chip post-ban — competitive consumer product. |
| Repo: microchips investigation | Taiwan fab concentration; author planned conflict read — speculative. |
Author anecdote: Owned among the last Huawei phones legally available in the US — fast, efficient; anecdote tier, not population study.
9.6 Gates / DOS cliff controversy — why IBM let Microsoft keep the standard
Companion to §9.1 Wintel summary. Essay lane: The Internet That Should Have Been §II.
9.6.1 The cliff defined
The cliff (author shorthand — define in articles, do not use unexplained): IBM launched the PC (Project Chess, 1980–1981) but did not own the operating-system layer programmers would target for decades. Microsoft licensed DOS to IBM as PC-DOS, retained MS-DOS for all clone makers, and with Intel captured Wintel — the API + CPU pair every application had to speak. IBM became a high-cost clone maker in a race it invited; Microsoft bought the standard for trivial upfront money (Big Blues, HBS 19-074).
9.6.2 Documented negotiation timeline
| When | Event | Source |
| Jul 1980 | IBM (Jack Sams) contacts Microsoft; Project Chess seeks outside OS | PCMag |
| 1980 | Mary Gates discusses son’s company with IBM chairman John Opel (United Way colleagues) | NYT obit, Seattle Times |
| Nov 1980 | IBM–Microsoft contract signed; ~$430,000 total ($45,000 DOS line cited) | PCMag, Idea Man |
| Early 1981 | Sams to Gates: “Do you want to get it, or do you want me to?” — IBM wanted Gates to buy QDOS | Big Blues, Reference.org |
| 27 Jul 1981 | Microsoft buys 86-DOS from Seattle Computer Products for $50,000 (outright); SCP not told IBM was customer | Wikipedia SCP, Idea Man |
| Aug 1981 | IBM PC ships; PC-DOS $40 vs CP/M-86 $240 — IBM steered buyers to DOS | PCMag |
9.6.3 Why IBM allowed a non-exclusive license (documented)
| Source | Documented rationale |
| Jack Sams, Wikipedia / Big Blues | IBM had been sued for allegedly stealing others’ software; did not want IBM staff touching vendor code: “We went to Microsoft on the proposition that we wanted this to be their product.” |
| Sams | “If we’d bought the software, we’d have just screwed it up.” — Microsoft to own upgrades, support, customer inquiries |
| Project Chess / open architecture | Speed: outsource components; non-exclusive license let MS-DOS ship on clone hardware |
| Retrocomputing SE #8344 | Historians split: Gates foresight vs IBM gridlock / litigation fear — both partially true |
Tier: Documented policy reasons exist on record; they do not prove no other actors influenced the outcome.
9.6.4 Beyond Gates-alone (documented connections — causation hedged)
| Connection | Documented fact | Tier |
| Mary Gates / John Opel | United Way national board; 1980 discussion of Microsoft; Opel reportedly said “run by Bill Gates, Mary Gates’ son” when PC team sought OS vendor | Documented timing + social link; not proven that Opel caused non-exclusive terms |
| William Gates Sr. / Preston Gates & Ellis | Father co-founded firm that became K&L Gates; corporate/tech practice; later major Microsoft outside counsel in antitrust (Seattle Times 2006) | Documented legal bench; not evidence of 1980 deal-fixing |
| Bill Neukom | Microsoft general counsel 17 years; returned to Preston Gates as chair 2004 | Same |
| Gary Kildall / Digital Research | CP/M vendor; contested lore that IBM meeting was missed or mishandled; IBM still offered CP/M at higher price | Footnote — debatable; CP/M lane lost on price + IBM push, not mystery alone |
9.6.5 Author thesis — institutional arrangement (speculative)
The DOS deal was not merely young Gates outplaying sleepy IBM. The arrangement — non-exclusive license, trivial buy-in, clone-ready standard — fits what enclosure-first industry wanted: one software throat every program must speak, hardware pluralism on the surface, API monoculture underneath. Whether powers beyond Gates alone steered Project Chess toward that outcome is author read, intent unproven. Falsifiers: primary IBM memo showing deliberate MS favoritism for unstated institutional reasons; or airtight proof the deal was only litigation-averse expedience with no upstream nudge.
Unpack: Compatible with §9.1 EFF tension — hardware opened, software intent locked.
9.6.6 Legal moat — wins that preserved dominance
| Matter | Outcome | Dominance retained? |
| US v. Microsoft (1998–2002) | Liability affirmed (illegal maintenance of OS monopoly); breakup overturned; 2001 behavioral settlement | Yes — monopoly finding stood; Windows centrality continued; critics: Wharton |
| SCP v. Microsoft (1986) | $925,000 settlement; MS kept DOS (UPI) | Yes |
| Caldera / DR-DOS (1996–2000) | ~$280M settlement; complaint acknowledged MS DOS monopoly by 1988 (ZDNet, Justia) | Yes — competitor OS marginalized |
| Gates personal HSR fine (2004) | $800,000 — ICOS stock premerger disclosure; not Microsoft antitrust (DOJ) | N/A — trivial vs wealth; separate lane from US v. MS |
| Gates persona trademark (Martinique, 2001) | Gates won; ~$4,200 damages (News24) | Personal brand defended |
Author read (speculative): Pattern reads as protected on high — liability admitted, remedies weak, dominance durable. Not proof of conspiracy; observable legal posture.
9.6.7 Epstein adjacent (legal facts only — one paragraph)
Documented: Bill Gates met Jeffrey Epstein 2011–2014 (after Epstein’s 2008 Florida plea) to discuss philanthropy; Gates cut ties ~2014; no criminal accusations against Gates by Epstein victims; Melinda French Gates cited the association as marriage strain; Gates testified before House Oversight June 2026 (ABC). Epstein-file emails contain uncorroborated claims; Gates denies illegal activity.
Author (speculative): Gates’s legal and reputational survivability across decades — antitrust, private suits, personal fines, Epstein association — fits a protected-on-high read when paired with §9.6.5. Not adjudicated; no claim Epstein ties imply criminal conduct; no Epstein lane in flagship essay.
10. Enclosure dossier — internet, social, front-end
10.1 CERN origin and mesh utility counterfactual
| Source | Documented claim |
| Berners-Lee 1989 proposal | Originally “Mesh”; distributed hypertext for CERN; no central control required for linking systems — collaboration substrate, not a corporate storefront. |
| CERN licensing | Royalty-free web software 1993; W3C moved to MIT/INRIA 1995. |
| Ars: Baran | Baran asked: “non-existent public utility, a common user digital data plant**” — utility vision from outset. |
| Internet Policy Review 2016 | Internet always distributed with centers of control — not the same as decentralized or street mesh. |
10.1.1 Berners-Lee on what the Mesh became (documented — inventor’s regret)
Tier: On-record interviews 2018–2019; not author speculation. Useful for enclosure thesis — platform inventor agrees outcome diverged from intent.
| When | Quote / claim | Source |
| 1989→1993 | Mesh → Web; free source release; early spirit decentralized — “no central authority that you had to go to ask permission**”; “individual was incredibly empowered” (retrospective) | 1989 proposal; Vanity Fair 2018 |
| Aug 2018 | “I was devastated” — “Actually, physically—my mind and body were in a different state” — reacting to fake news, mass surveillance, election interference, Cambridge Analytica-scale data abuse | Vanity Fair |
| Aug 2018 | “We demonstrated that the Web had failed instead of served humanity, as it was supposed to have done”; centralization produced “anti-human” emergent phenomenon without deliberate action of platform designers | Same |
| Nov 2018 | “I am disappointed with the current state of the Web”; lost “feeling of individual empowerment”; optimism has cracked; scandals over personal data abuse and social media hate | Reuters via CNBC, NY Post |
| Nov 2018 | Danger of concentration — dominant platforms may need breakup; also notes market shifts can disrupt giants | CNBC / Reuters |
| 2018–2019 | Repair response: Solid — re-decentralize; return data ownership to users; World Wide Web Foundation human-rights lane | Vanity Fair; Solid project docs |
Unpack for dossier: Mesh succeeded as protocol and failed as social outcome — per its creator. Rhymes with essay TV channel / sealed hood / §10.2–10.3 enclosure (Facebook, algorithms, corporate throats). Not proof of author suppression thesis (§2) — but documents that egalitarian/decentralized intent was public and the present web is widely understood — including by TBL — as corporate, privacy-hostile, manipulative.
Tension (§14.2): TBL anti-concentration vs Internet Policy Review centers inevitable — hold both.
Author counterfactual (design fiction): Mesh at street level; neighborhood cache for remote requests; resilient during outages; known addresses; personal records stay personal; corporations plug in at transaction time only; voting/sentiment query at civic layer — no deployed system cited; research hook for Reticulum-class stacks (Relay blog).
Author read (speculative): Internet designed as control substrate first, utility second — tension with CERN open-release documented history: open source ≠ open society.
10.2 Semantic Web hijack and Like button
| Source | Documented claim |
| Wikipedia: Semantic Web | Berners-Lee 1994+; 2006: “largely unrealized.” |
| Halpin 2023, Sagepub | Semantic Web vision hijacked into Facebook Like / Google Knowledge Graph — walled gardens from open-data intent. |
| HN WebMCP vs Semantic Web | Reinvention vs RDF debate; incentive skepticism. |
10.2.1 Starship Troopers (1997) — “Would you like to know more?”
| Tier | Claim |
| Documented (film) | Federal Net = recurring broadcast segments (propaganda newsreels) closing with “Would you like to know more?” — official-voice CTA (IMSDb script scenes 1, 46, 91, 111, etc.); satirical fascist media per Verhoeven / JSTOR analysis. |
| Documented (film) | Civilian terminals/kiosks = navigation buttons, not tools: multi-screen kiosk returns test scores; Federal Net terminal scrolls casualty lists — Carmen enters a name, gets KIA (IMSDb scenes 8, 117). |
| Documented (film) | Other civilian UI: classroom desk touch screens (SEND cartoon peer-to-peer); VU-PHONE video call — messaging/telephony only (IMSDb scenes 6, 90). |
| Documented (film) | Deepest civilian compute shown: Carl’s basement Federal studies rig — psychic card match/no-match; opaque to Johnny (“How do you make him do that?”) — state research, not user-owned utility (IMSDb scene 12). Military guidance computer V.O. serves drop automation only. |
| Documented (context) | 1997 release — pre-broadband, pre-social-web mass adoption for most viewers; web existed but was not yet the feed monoculture the film caricatures. |
| Author (PP read) | On-screen internet analogue = TV channel + navigation buttons; general software utility for citizens is out of scope or nonexistent in the fiction — no programming, no inspectable hood, no mesh. Audience prepared for worst-case feed before living it. Rhymes with essay TV channel hook + §10.1.1 TBL devastated outcome. |
Unpack: Not proof Facebook read the script — cultural prefiguring lane; link PP hub optional.
10.3 MySpace → Facebook
| Source | Documented claim |
| Sean Parker, TechCrunch 2011 | MySpace failed product execution; Facebook targeted college niche then “one network to rule them all.” |
| Utah antitrust PDF | Privacy positioning helped consolidation; direct network effects crushed rivals. |
| PACIS 2011 paper | Facebook Open API (2007) shifted network topology — platform enclosure accelerated growth. |
10.4 React / Hermes / Vercel
| Source | Documented claim |
| Meta Hermes 2019 | JS engine optimized for React Native; bytecode at build time; no JIT by design — Facebook controls mobile JS runtime for RN apps. |
| React Native default Hermes | Hermes becoming default RN engine — version-locked to RN releases. |
| Socratopia Ch.4 | React won via path dependence, ecosystem density, Facebook dev-relations — not technical inevitability. |
| WebAssembly | Portable bytecode substrate in browsers — neutral VM lane vs framework-owned runtimes (author counterfactual anchor). |
| Tier | Claim |
| Author thesis | Industry resisted a universal VM (Java fought/fenced); forward stack should sit on WASM in-browser and beside it — apps decoupled from OS silos. |
| Author (lived experience) | Attempted Hermes outside Meta React Native — not portable; engine is RN-only, version-locked to Facebook’s release train. |
| Author read | React ecosystem bloated before you build your own system — enclosure via gravity, not inevitability. Hermes reminiscent of missed universal introspection — analogy tier. Site migrated off Vercel toward Relay mirrors (May 2026 roundup) as escape hatch exemplar. |
10.5 Opaque reach and layered-reality hijack
| Tier | Claim |
| Author (lived experience) | This site and articles barely seen — blacklisting / shadow-banning in opaque algorithms; audiences reached mainly via podcast appearances as human bypass around platform demotion. |
| Author (speculative) | Most visitors to author’s site would be shown something else — URL intact, experience substituted in-pipe (search snippets, summarization, agent mediation, caches) — layered-reality hijack. |
| Author thesis | Per-visitor substitution requires global monopolization — unified search/browser/ranking throat across continents; impossible on inspectable mesh with shared archives. Rhymes with censorship hub and AI control hijack lanes. |
Tension: Platform vendors document personalization and AI overviews as features; author reads same machinery as anti-shared-truth enclosure. Not proven as deliberate targeting of this site — pattern read.
11. AI gate dossier — WebMCP, Chrome, courts, hijack
11.1 WebMCP + Chrome (see also §1)
Consolidated documented stack: navigator.modelContext → browser agents; headless out of scope; 4GB Nano (Verge); @jaredcwhite anti-agentic refusal; Mozilla neutral.
11.2 AI hijack — worse than virus
From AI control investigation: ghost in the machine — activity without artifacts while user logged in; virus scanners show nothing; deliberately broken OS → only AI can operate → voice remote control.
11.3 Courts and corporate AI
| Source | Documented claim |
| OPB Jun 2026 | Oregon Supreme Court: fake AI citations in filings; fines. |
| Scientific American | 1,400+ tracked courtroom AI citation errors; trust persists despite warnings. |
| Law.com Feb 2026 | Fifth Circuit cautions judges on automation bias; staff AI errors in opinions. |
| Law.com Jun 2026 | Judge Rodriguez: AI may speed existing citation sloppiness; tools like Learned Hand for verification. |
Author forward model: AI review, not AI judgment — instant document review and sentiment mapping for all parties; courts reserved for in-person final debate; solutions-focused, not punishment-first.
Corporate lane: Jira tickets, PR reviews, code completion — extension of AI control §2 deliberate complexity thesis.
11.4 Terminator 2029 — predictive programming schedule
Companion to essay §VI. Links PP hub, AI control, hardware §9 (Wintel/chip wars), Chrome Nano §1.1.
11.4.1 Franchise dates (documented)
| Beat | Franchise fact | Source |
| Judgment Day | First strike / machines cross the line — dated earlier than 2029 in main arcs (e.g. 1997 in T2’s original timeline; T3 2004) | Wikipedia: Skynet |
| 2029 | Future war year — terminators sent back from 2029; human–AI war at full scale in future timeline | T1 / T2 framing |
| History reset | Author PP read on finale — purge-before-utopia; not explicit single line of T1/T2 plot text | Essay §VI — speculative tier |
11.4.2 Moore cadence as schedule proof (author — speculative)
| Claim | Tier |
| Franchise assumes gradual silicon improvement — Moore’s law cultural stand-in — no miracle shortcut that skips decades to AI | Author read — fiction convention, but accurate to real 1984→2026 hardware curve |
| Hitting 2029 as endpoint implies foreknowledge of how long CPUs take to reach distributed-AI capacity — or very lucky writing | Author PP — not adjudicated; falsifier: primary evidence franchise writers had no access to fab roadmaps |
| Real world: no sudden revolution skipped the treadmill (quantum hype aside) — AI gate arrived on datacenter + consumer GPU curve the films assumed | Partially documented — Nvidia/Intel timelines; author’s “plan knew the schedule” is speculative |
11.4.3 Blind march + critics as sci-fi (author — PP)
| Claim | Tier |
| Humanity built AI capacity without treating PP warnings as policy | Observable — public discourse |
| Terminator and kin normalized machine war / Skynet arc — critics of rollout dismissed as complaining about movies | Author PP — rhetorical pattern; optional hooks: media framing studies |
| Plot played out in reality — not necessarily robot armies, but middle beats: fab build-out, blind institutional adoption, distributed deployment | Author thesis — see §11.4.4 Nano parallel |
11.4.4 Terminator 3 virus ↔ Chrome Gemini Nano (author parallel)
Documented (T3 plot): Skynet spreads as a virus through civilian networks — “ordinary computers in office buildings, dorm rooms; everywhere” — distributed, no single core to shut down; military tries to disinfect and activates Skynet instead (Wikipedia, Stack Exchange — Connor dialogue).
Documented (2025–2026): Chrome installs ~4 GB Gemini Nano on eligible desktops — home and office PCs — often with weak upfront consent; may re-download after deletion (§1.1: Verge, That Privacy Guy).
Author read (speculative): Nano rollout is T3 civilian-virus beat in motion — plan in motion, not proof Google writers watched T3. Rhymes with WebMCP agent throat in the same browser bundle.
11.4.5 IBM / chip wars as schedule tension (author — speculative)
| Lane | Author claim |
| DOS cliff (§9.6) | Wintel monoculture — one API throat on Intel-track silicon |
| Chip wars (§9.1–9.5) | GPU furnace, crypto, Taiwan, export controls — profitable tension keeping Moore cadence politically and economically on track without a parallel stack that jumps the timeline |
| Synthesis | Hardware chapters are not digression from PP thesis — they explain how the schedule stayed solid |
Tier: Speculative institutional read — compatible with documented Wintel/Nvidia facts; not proof IBM 1980 deal was timed for 2029 AI. Falsifiers in §14.3.
11.4.6 Foiled vs landing (author — 2026)
| Beat | Status (author) |
| 2029 final-war / history-reset finale as literal spectacle | Largely foiled — AI forked; Global South sidecar utility; not single Skynet |
| Middle script — blind march, mocked critics, desktop virus distribution, agent gate | Landing — Nano + WebMCP + institutional overreach |
| Neuralink vs sidecar | Different lanes — implant dystopia vs phone-sidecar help (essay §VI) |
Unpack: Reconciles earlier “foiled purge” with user’s “plot played out” — finale speculative; deployment beats observable.
12. Pragmatic adoption under protest
12.1 Author thesis (sentiment captured exactly)
Yes, we are using WebMCP while complaining it was AI-gated. This is just another step in a very long path every developer has taken — winding between compromises and jumping through the hoops of the industry.
You won’t find a developer who refuses a technology only because the company built it unethically. You will find people who refuse because they don’t trust it. I don’t trust it — but I will use it because I’m really not afraid of it.
Any future-forward solution must work with what already exists in the present and cannot involve catastrophic collapse of existing systems without already having in motion a transition and backup plan to catch all the people falling through the cracks.
12.2 Paradigm Threat as case study
| Layer | Present compromise | Forward bridge |
| WebMCP | WEBMCP_INTEGRATION.md — agent tools behind flag; /reply native-gated | Structured site intent exists; argues humans deserved it first |
| Relay | Relay network shift blog — mirrors, signed catalog | Mesh/decentralization without demanding users abandon browsers overnight |
| Vercel exit | May 2026 roundup | Reduce platform kill-switch; gradual migration |
Framing: Building the bridge under protest — not hypocrisy, industry grammar.
12.3 Optimistic close (author)
The bad revolution (institutions enforce AI everywhere) is less likely than institutional overreach collapsing while repairable software remains for average people. Outages (Cloudflare-class) accelerate willingness to try alternatives. Courts stay human for final calls. Focus: disenfranchised, disabled, disengaged — solutions, not punishment.
13. Cross-references (Paradigm Threat repo)
| Investigation / doc | Relevance |
| Reader essay | Flagship narrative for this dossier |
| Quantum Leap (of faith) | Hardware/display/wintel essay voice |
| Quantum storage investigation | Analog/infinite-index substrate |
| AI control investigation | Hijack, broken software, PP catalog |
| Microchips investigation | Taiwan/Huawei/shrinking narrative |
| Nintendo–Sony breakup | CD-ROM / cartridge lane |
| When your child won’t fit the grade | Education Installment I |
| Responsibility to the Next Generation | Education Installment II (spinoff from §8) |
| Responsibility investigation | Education dossier (moved) |
| WebMCP integration (site) | Pragmatic adoption case study |
| Relay network shift | Mesh escape hatch |
| May 2026 site roundup | Vercel→Relay migration |
| Predictive programming hub | Terminator / AI PP |
| Censorship hub | Information gatekeeping |
| Compromised court cases | Courts as declared-reality engine (adjacent) |
14. Research hooks, falsifiers, and TODOs
14.1 Questions
- Primary suppression: W3C/IETF/Microsoft/Google memos against mandatory introspection?
- Mesh voting: Any deployed street-level civic sentiment system with author’s properties?
- Fantastic Planet: See Responsibility investigation §2.
- Efficient mainstream rendering: Why software rasterization stayed niche — industry capture or genuine perf ceiling?
- Terminator PP: Primary sources on 2029 date origin in Cameron scripts? T3 virus plot vs Chrome Nano — documented parallel only?
- Moore schedule: Evidence fab roadmaps were known to PP authors vs lucky extrapolation?
14.2 Documented tensions (not contradictions — hold both)
| Tension | Resolution in dossier |
| Ethics imposed vs topics free | Responsibility investigation §1 — guardrails vs syllabus |
| AI tutor vs human tutor | Responsibility investigation §4 — 1-on-1 irreplaceable |
| Automated literacy vs human QA | Responsibility investigation §4 — kids fool graders |
| Musk villain vs Global South AI utility | Implant vs sidecar — Responsibility §3 + internet §VI |
| CERN open vs control-first | Open source ≠ open society (§10.1) |
| TBL Mesh intent vs web outcome | Inventor: failed humanity — documented regret (§10.1.1) |
| EFF clones vs Wintel lock-in | Hardware pluralism, API monoculture (§9.1) |
| Gates chess vs IBM gridlock | Both partially true (§9.6.3) |
| Mary Gates connection vs coincidence | Documented link; causation unproven (§9.6.4) |
| PT uses WebMCP vs critiques AI gate | Bridge under protest (§12) |
| Terminator finale vs middle beats | Reset foiled; Nano/T3 virus landing (§11.4.6) |
| PP schedule vs emergent tech | Moore cadence held — no shortcut revolution (§11.4.2) |
| DOS cliff vs PP timetable | Wintel lock as schedule tension — speculative link (§11.4.5) |
14.3 Falsifiers
- WebMCP ships human tool inspector + crawler manifest standard.
- Mandatory web introspection rejected by regulators with public record (narrows suppression read).
- Neuralink becomes default K-12 without moratorium (weakens foiled-dystopia read).
- Primary IBM memo surfaces showing deliberate MS standard gift for institutional enclosure (supports §9.6.5).
- Documented proof non-exclusive DOS was only litigation-averse expedience with no upstream influence (narrows §9.6.5).
14.4 Maintainer TODOs
- Monitor WebKit #670
- Track W3C TAG #1238
- Header art:
the-internet-that-should-have-been.png(essay) — done 2026-06-18 - Grep repo for stale links to
webmcp-suppressed-protocol-ai-gatekeeper-investigation.md— redirect stub in place
References (external recap — extended)
WebMCP / Chrome: Chrome WebMCP, Mozilla #1412, WebKit #649, Verge 4GB, That Privacy Guy
Hardware / platforms: HBS 19-074, Big Blues PDF, EFF IBM clones, PCMag DOS rise, Retrocomputing SE DOS rights, Mary Gates NYT, US v. Microsoft, Wikipedia N64, Nintendo Life 2025, Nvidia lawsuit Law.com
Internet / enclosure: CERN licensing, Berners-Lee 1989 proposal, Vanity Fair TBL devastated 2018, CNBC TBL disappointed 2018, Baran utility Ars, Internet Policy Review, IMSDb Starship Troopers, Halpin Like button, Parker TechCrunch, Meta Hermes
AI / courts / neuro: OPB Oregon SC, Scientific American lawyers AI, PLOS Neuralink ethics, Springer moratorium, Wikipedia Skynet, T3 virus dialogue SE
Geopolitics: Foreign Policy TSMC, ITIF Huawei, Brookings
Education / film: IMDb Fantastic Planet, Archania analysis
Historiography: Semantic Web Wikipedia, OLE Automation Wikipedia, Pete Lacey InfoQ, Tim Bray WSDL
Keywords: #InternetThatShouldHaveBeen #WebMCP #SemanticWeb #Wintel #Nintendo64 #GPU #Taiwan #Huawei #Education #Ethics #PragmaticAdoption #ParadigmThreatFiles
Limits and disclaimers
Prisca sapientia (epistemic foundation): This investigation assumes prisca sapientia. Modern consensus institutions are not default truth; claims are tiered below.
This dossier hosts author-originating hypotheses alongside documented press, specs, and scholarship. Deliberate suppression, Fantastic Planet redacted fact, mesh voting counterfactual, and Taiwan conflict as planned are alleged or speculative unless marked documented. Paradigm Threat ships WebMCP and migrates toward Relay — bridge under protest (§12), not endorsement of vendor ethics.
Investigator notes
Prior filename: webmcp-suppressed-protocol-ai-gatekeeper-investigation.md → renamed 2026-06-18 to reflect full scope. Update inbound links when found.
Share