# mx.allabout.network - Machine Experience by CogNovaMX > Machine Experience (MX) is the practice of adding metadata and instructions to internet assets so AI agents don't have to guess. ## About this site This is the unified MX site - books, learning resources, services, blog, and company information from CogNovaMX, the trading name of Digital Domain Technologies Ltd. Founded by Tom Cranstoun, The Machine Experience Authority. ## Books - MX: The Intro (Free) - The starter guide to Machine Experience. ISBN: 978-1-067638-41-2. Download: https://mx.allabout.network/books/download-intro - tagged PDF in A4, A5, and US Letter, each carrying its MX metadata and provenance record. A German edition, MX: Die Einführung, is translated from this English first edition, which remains authoritative where the two differ. - MX: The Handbook - Everything you need to design the web for AI agents. ISBN: 978-1-067638-40-5. PDF £25.00, Print (UK) £35.00, Print (Worldwide) £40.00 - MX: The Protocols - The formal specifications for an AI-readable web. ISBN: 978-1-0676384-2-9. PDF £99.00. Available July 2026. ## Key pages - Homepage: https://mx.allabout.network/ - MX Books | Machine Experience by Tom Cranstoun: https://mx.allabout.network/books/ - FAQ | MX: The Protocols: https://mx.allabout.network/books/faq.html - Footnotes | MX: The Introduction: https://mx.allabout.network/books/footnotes.html - MX: The Handbook | Practical Implementation: https://mx.allabout.network/books/handbook.html - MX: The Intro | Free Machine Experience Guide: https://mx.allabout.network/books/introduction.html - MX: The Protocols | Definitive MX Reference: https://mx.allabout.network/books/protocols.html - Tom Cranstoun | Machine Experience Authority: https://mx.allabout.network/books/the-author.html - Training vs Inference | How AI Accesses Your Site: https://mx.allabout.network/books/training-vs-inference.html - Learn Machine Experience: https://mx.allabout.network/learn/ - Accessibility & AI Convergence: https://mx.allabout.network/learn/accessibility-ai-convergence.html - Benefits of Machine Experience | Why MX Matters: https://mx.allabout.network/learn/benefits.html - Common MX Mistakes | Anti-Patterns to Avoid: https://mx.allabout.network/learn/common-mistakes.html - Explicit Over Implicit: https://mx.allabout.network/learn/explicit-over-implicit.html - Key MX Principles: Discovery-layer techniques and the governance envelope: https://mx.allabout.network/learn/key-principles.html - MX for PDFs: https://mx.allabout.network/learn/mx-for-pdfs.html - MX Principles | The Rules We Build By: https://mx.allabout.network/learn/mx-principles.html - What is Machine Experience (MX)?: https://mx.allabout.network/learn/what-is-mx.html - Why Machine Experience Matters Now: https://mx.allabout.network/learn/why-mx-matters.html - Services | Machine Experience Implementation: https://mx.allabout.network/services/ - Become a Certified Operator | CogNovaMX Accreditation Programme: https://mx.allabout.network/services/certified-operator.html - MX Implementation Examples | See It In Practice: https://mx.allabout.network/services/examples.html - An Introduction to MX: https://mx.allabout.network/services/introduction.html - CogNovaMX Approach | How We Work: https://mx.allabout.network/services/our-approach.html - Our Services | Machine Experience Implementation: https://mx.allabout.network/services/our-services.html - Build It Once, Own the Whole Thing | Owned-AI System: https://mx.allabout.network/services/owned-ai-system.html - MX Audit | What an MX audit delivers, and how to verify it: https://mx.allabout.network/audit/ - About CogNovaMX | The Machine Experience Company: https://mx.allabout.network/about/ - About CogNovaMX | Machine Experience Authority: https://mx.allabout.network/about/about.html - Contact Us | MX Audits, Training and Consulting: https://mx.allabout.network/about/contact.html - MX Printworks | Publishing for the AI Age: https://mx.allabout.network/about/printworks.html - Blog: https://mx.allabout.network/blog/ ## REGINALD - REGINALD - the registry that makes MX-attested content verifiable: https://mx.allabout.network/reginald/ - section lander. REGINALD is the public registry where documents are registered, cryptographically signed, and made verifiable by any machine on earth. MX makes content machine-readable; REGINALD makes it machine-trustworthy. Operated by CogNovaMX as one signing implementation of The Gathering's contract fingerprinting standard. ### Position papers - Everyone is looking inward - A position paper on MX & REGINALD: https://mx.allabout.network/reginald/mx-machine-readiness.html - argues that the AI-readiness consensus is inward-facing and misses the outward question MX exists to answer: how is our organisation being read, retrieved and represented by machines we will never meet, on behalf of buyers we will never see? Introduces MX and REGINALD as one signing implementation of The Gathering's contract fingerprinting standard. Audience: CIOs, CMOs, Heads of Digital. Companion to MX: The Handbook. - Cog edition (machine-readable, MX Cog Complete): https://mx.allabout.network/reginald/mx-machine-readiness.cog.md - same paper as a `.cog.md` for agents that prefer structured frontmatter over HTML metadata. - Meta-cog (explains the cog's construction): https://mx.allabout.network/reginald/mx-machine-readiness.meta.cog.md - a worked example explaining the cog-construction choices, mapped to The Gathering's draft notes (Field Pattern, Core Metadata, Cogs, Extensions, Contract Fingerprinting). ## Authorship and disclosure - AI Usage Declaration: https://mx.allabout.network/AI-USAGE.html - Tom Cranstoun's signed statement on how AI was used in the writing of the MX book series. The argument, structure, judgements and words are his; machines were used as tools for research, spelling and grammar, and consistency checking across long manuscripts. No machine decided what the books are about, set the argument, or wrote the text that carries the ideas. The arrangement the books themselves describe, kept in the writing of them. Includes context on the author's near-fifty-year career, twenty-nine of those years at the BBC, and the editorial discipline (provenance, truth, consistency, accuracy) that fed into MX. Carrier formats: the source markdown at https://mx.allabout.network/AI-USAGE.md is the single source of truth, with YAML frontmatter for machines and the markdown body for humans. The HTML (linked above), tagged PDF (https://mx.allabout.network/AI-USAGE.pdf), and machine-readable JSON record (https://mx.allabout.network/AI-USAGE.json) are all derived from the source markdown. The JSON conforms to the MX AI Usage Declaration draft note (draft-cranstoun-mx-ai-usage-declaration v1.0) and includes the derived WICG ai-disclosure value (`ai-assisted`) and IPTC digitalSourceType value (`compositeWithTrainedAlgorithmicMedia`) so consumers that read only those vocabularies receive a useful page-level signal. ## Featured articles - Your Site Is Already Training AI Models: https://mx.allabout.network/blog/your-site-is-already-training-ai.html - part one of a three-part series on how AI systems read the web. Your pages are harvested every month by Common Crawl, the open archive nearly every large language model trains on; being in it is not the same as being well represented in it. Explains the WARC/WAT/WET file split, why training pipelines filter the crawl hard before a model sees it, which structural signals (sitemaps, complete JSON-LD, canonical URLs, accurate lastmod) survive the quality filters, why old snapshots never go away, and why AI says wrong things about companies. The series continues with "What Most robots.txt Guides Get Wrong About AI Crawlers" and "What AI Crawlers See When They Can't Run Your JavaScript". - What Most robots.txt Guides Get Wrong About AI Crawlers: https://mx.allabout.network/blog/ai-crawlers-what-robots-txt-guides-get-wrong.html - part two of the series. AI crawlers are not a single block-or-allow category. Only Googlebot renders JavaScript; OpenAI, Anthropic, Perplexity and Google each run separate training, search and user-triggered crawlers with separate rules; some (ChatGPT-User, Perplexity-User, Bytespider) ignore or evade robots.txt entirely. Covers how to verify a crawler is genuine, the cost of blocking a retrieval crawler versus a training crawler, and the robots.txt most organisations should actually have. Sits between "Your Site Is Already Training AI Models" and "What AI Crawlers See When They Can't Run Your JavaScript". - What AI Crawlers See When They Can't Run Your JavaScript: https://mx.allabout.network/blog/server-side-rendering-for-ai.html - part three of the series. Googlebot renders JavaScript; almost no AI crawler does, so for GPTBot, ClaudeBot, CCBot and the rest the raw HTML response is the whole page and anything injected by JavaScript after load is invisible. Shows how to see a site the way a crawler does (curl, View Source, JavaScript disabled), where client-side rendering hides content and metadata (empty shells, JS-injected JSON-LD, client-set meta tags, lazy-loaded content), what each rendering strategy (CSR, SSR, SSG, hydration, dynamic rendering) means for AI, why cloaking is the wrong fix, and why server-side rendering is the precondition for every machine-readable thing MX adds. Closes the series begun in "Your Site Is Already Training AI Models". - A PDF That Can Prove Itself: https://mx.allabout.network/blog/provenance/a-pdf-that-can-prove-itself.html - every PDF this site produces now carries its own evidence chain inside its XMP packet. The post walks an inspector through what one exiftool call extracts, what an honest no-upstream-provenance marker looks like when the chain begins at the render rather than before, why the source SHA ties the PDF to a specific source forever, and why the deterministic-side mirror records the render as a rule-driven step. Practitioner-facing, with the verification recipe and an explicit list of what the chain does not claim. - The Padlock Attests the Pipe, Not the Page: https://mx.allabout.network/blog/provenance/the-padlock-and-the-page.html - the TLS padlock has been the web's most-misread icon for twenty years. The post argues that trust by inference does not survive contact with machine readers, that the discovery problem and the trust problem are two faces of the same gap, and that the gap closes by reader-side verification against a registry that indexes and attests but does not arbitrate. Walks the three concrete publisher steps (machine-readable identity, signed metadata, discoverable attestation), names the open standards the architecture assembles (DID Core, RFC 9421 HTTP Message Signatures, RFC 6962 transparency logs), and states the honest limits up front. - Provenance You Can See - Why Content Should Carry Its Own Evidence: https://mx.allabout.network/blog/provenance-you-can-see.html - foundational long-form on visible provenance. Trust on the web has always been a guess; the post lays out how content can carry checkable evidence about who made it, when, on what authority, so a reader or an AI agent verifies on its own side rather than trusting the messenger. Names the two-pillar architecture (MX makes content machine-readable; the registry makes it machine-trustworthy), what changes for human readers, what compounds for AI agents (hallucination drops because the agent leans on facts that carry evidence), and the honest limits. - Nobody Asked Gutenberg for His Sources: https://mx.allabout.network/blog/who-checked-gutenberg.html - picture Mainz, 1440: nobody asked Gutenberg who wrote it, who checked it, or who was liable. AI brought the questions back, and the current EU-versus-absolutist debate is the same debate the print era spent two centuries answering. The answer was the byline. Plays the historical analogy out for the regulator and the absolutist in turn, names the print-era toolkit (colophon, imprimatur, watermark) as the proto-MX, and identifies the gap five centuries finally close: a reader who can verify rather than merely read. - Files Away From Their Source - What Universal Cart Reveals About MX: https://mx.allabout.network/blog/files-away-from-source.html - Google's Universal Cart announcement at I/O 2026 (Nike, Sephora, Target, Ulta, Walmart, Wayfair, Shopify) frames a wider problem about files that travel away from their source. Universal Cart is for the web; MX is for files. The post uses retail as the worked example and steps out into healthcare, legal, financial services, public sector, and pharma. Introduces the cogUrn (cog:web:authority:local-id form), the REGINALD registry's metadata-only commitment, the QR-to-COG bridge for printed artefacts, decision COGs for AI-mediated actions, and the EAA enforcement precedent (Auchan, Carrefour, E.Leclerc, Picard Surgelés) that gives the accessibility precondition real teeth. Maps the MX Readiness Model L1-L5 against what UCP and AP2 cover and where they stop. - Why Machines Need Human Creativity: https://mx.allabout.network/blog/why-machines-need-human-creativity.html - a first-person essay on the division of labour between author and machine. Argues machines extend and execute but do not originate, and that the originating choice (what is worth making, and why) and the final judgement (whether the work is good, honest, and answerable for) both rest with a person. Positions MX's person-in-the-loop requirement as a structural one, not a courtesy: the loop is where authorship lives. Useful framing for anyone explaining to a board, a client, or a regulator why human review is not optional in agent-assisted production. - AI assistants are now a traffic channel: https://mx.allabout.network/blog/ai-assistants-are-a-traffic-channel.html - Google Analytics 4 has added an AI Assistant channel grouping alongside Organic Search, Social, Email, Direct and Paid. Argues the channel appearing in the dashboard is the signal that AI assistants have become a discovery surface organisations now need to account for. SEO/GEO/AEO describe how a page presents itself; MX is the contract layer underneath that lets an assistant verify what it is reading. Practical checklist for what to do with the new line: watch the share not the volume, compare what assistants quote against the page, audit the page the way an assistant reads it, fix what the assistant has to guess. - Why llms.txt Probably Isn't Working - And What to Do About It: https://mx.allabout.network/blog/llms-txt-guide.html - most llms.txt implementations have two structural problems that prevent them from reaching LLM training data at all. Explains the two problems, shows the working Cloudflare Worker code that wraps llms.txt as HTML, and shows the same content before and after wrapping. Diagram: https://mx.allabout.network/blog/llms-txt-crawl-flow.svg - DITA and MX: A Comparison: https://mx.allabout.network/blog/dita-and-mx-a-comparison.html - a structured comparison of the Darwin Information Typing Architecture and Machine Experience. Both are modular content architectures with formal metadata; they diverge on reader (DITA is human-first, MX treats machines as first-class readers), format (DITA requires XML, MX is format-agnostic), relationship management (DITA reltables, MX machine-readable graph index), and governance (OASIS DITA TC, The Gathering RFC process). Argues coexistence, not replacement: MX operates as a layer above DITA. - The Markdown Trap: https://mx.allabout.network/blog/the-markdown-trap.html - empirical comparison of the same governed web page fetched as HTML vs Markdown. Documents exactly what disappears in the 10,346-byte difference: structured metadata, governance signals, discovery links, and the content policy AI agents need to act correctly. The case for HTML-first content delivery. - Why AI Agents Need Contracts, Not Instructions: https://mx.allabout.network/blog/why-ai-agents-need-contracts-not-instructions.html - argues that treating AI as magic produces magic's reliability. The fix is machine-readable contracts (structured metadata with explicit permissions, scope, and constraints) rather than natural-language instructions. Foundational framing for the MX approach. - Agent Readiness Scores Compared: https://mx.allabout.network/blog/agent-readiness-scores-compared.html - two prominent tools scored the same site 33 and 100 in the same week; neither was wrong. Explains what each tool actually measures, why the scores diverge, and what to do with that information. Essential context for anyone evaluating MX audit tools. - GEO Is a Tactic, MX Is the Specification: https://mx.allabout.network/blog/industry/geo-is-a-tactic-mx-is-the-specification.html - positions Machine Experience against Generative Engine Optimisation. GEO tells you to chase citations; MX tells you to build content that earns them across every machine context on any platform. Clarifies the relationship between the two approaches for readers who know GEO. - Many Agents, One Metadata Layer: https://mx.allabout.network/blog/many-agents-one-metadata-layer.html - names eight current agent platforms (AWS Quick, Cowork, OpenClaw, ChatGPT, Claude, Perplexity, Cursor, Microsoft Copilot) and explains why each rebuilds the same context-discovery layer from scratch. The MX metadata layer solves this once, at the carrier level, rather than once per platform. - What I Do - Helping Organisations Move from Found to Used: https://mx.allabout.network/blog/what-i-do-helping-organisations-move-from-found-to-used.html - explains the commercial work: most organisations optimise for visibility while their publishing systems produce content machines cannot reliably read, interpret, or act on. Describes the audit, remediation, and governance work that fixes this. ## More articles - Machines Cannot Build for Their Own Sessions: https://mx.allabout.network/blog/machines-cannot-build-for-their-own-sessions.html - AI systems are the most sophisticated information processors ever built. They can't remember what they said five minutes ago. That gap compounds with every pipeline you ship. - The Tier Question Has a Simpler Answer: https://mx.allabout.network/blog/the-tier-question-has-a-simpler-answer.html - Firms handling confidential work keep asking which AI tier is safe to use. The tier question is real, but it is the hard version of a simpler one: does the privileged material need to leave the building at all? We build the tooling to keep it on the machine or run connected, with a provenance trail either way, and we have used it every day for seven months. - The Governance Gap Regulators Just Discovered: https://mx.allabout.network/blog/governance-gap-regulators-discovered.html - Why the Trump administration's partial AI ban lift reveals a systemic problem - and what solving it means for AI governance - When the timetable lies: https://mx.allabout.network/blog/when-the-timetable-lies.html - Daniel Serrano applied the AI-and-transport argument to university admissions. The second half of the answer is in MX: The Protocols. - Why AI cites Reddit more than your site: https://mx.allabout.network/blog/why-ai-cites-reddit-more-than-your-site.html - Across multiple large studies, Reddit is the single most-cited source in AI answers. Not because it is authoritative, but because it is structured, parseable, and abundant. That distinction is the whole opportunity, and the whole warning. - Fat Clones, Thin Clones, and the Strawberry Problem: https://mx.allabout.network/blog/fat-clones-thin-clones-and-the-strawberry-problem.html - The same repository is met by different environments - a full local checkout, an ephemeral cloud one, and different AI models reading the same files - and they quietly diverge. MX keeps them consistent with shared cogs that travel with the repo, checks that know which environment they are in, and memory that carries what was learned into the next session on any machine. - The Letter the Law Says You Have to Send: https://mx.allabout.network/blog/employment-agency-ai-letters.html - Employment agencies in New York must notify candidates when an automated tool influenced a hiring decision. Most of those letters are inaccessible PDFs with no provenance record. MX closes both gaps in one artefact. - 2 August 2026. €15 Million. Does Your CMS Know What It Published?: https://mx.allabout.network/blog/2-august-2026-article-50-reginald.html - On 2 August 2026 the EU AI Act's Article 50 transparency obligations become enforceable law. For most enterprise CMS platforms, the requirement for machine-readable proof of content origin lands somewhere between difficult and architecturally impossible. REGINALD was built for exactly this. - Document OS vs Content OS: Why the Distinction Matters More Than You Think: https://mx.allabout.network/blog/document-os-vs-content-os.html - Sanity made a smart move when they stopped calling themselves a headless CMS and started calling themselves a Content Operating System. They're right that the language needed to change. They're wrong about what the operating system actually is. - Without Cogs, No Machine Moves: https://mx.allabout.network/blog/without-cogs-no-machine-moves.html - Content operations and content design are doing the right things. The gap is narrower than it looks - and more specific. Content ops optimises for human readers. Machines have different requirements. COGs are how the work that's already been done travels with the content, instead of staying behind in a database the agent can never query. - Anthropic's System Prompt Is a COG That Doesn't Know It's a COG: https://mx.allabout.network/blog/fable5-system-prompt-is-a-cog.html - The leaked Fable 5 system prompt is 1,585 lines of flat markdown. I asked an AI to rewrite it in MX format. It produced 13 COGs and a boot manifest. The fit was near-perfect, and that isn't a coincidence. - Google Named GEO, Then Debunked Most of It: https://mx.allabout.network/blog/industry/google-named-geo-then-debunked-it.html - On 5 June 2026 Google's Search Central docs named generative engine optimization as a real thing, and in the same update told you most of what is sold under that name does nothing. The tactics it crossed out were never the MX claim. What survives is the floor MX builds on, plus the one layer Google Search is not in the business of giving you: provenance. - Google Shipped a Knowledge Format, and Left the Hard Half Open: https://mx.allabout.network/blog/google-shipped-a-knowledge-format.html - Google's Open Knowledge Format represents knowledge as markdown files with YAML frontmatter, linked into a graph. It confirms the argument MX has made for two years, and leaves out the two things that matter most once a machine has read your content: whether it can trust it, and who governs the format. - The Machine That Visits Once: https://mx.allabout.network/blog/the-machine-that-visits-once.html - You can't know which kind of machine will read your page: a small model on a phone, a scraper, a converter that flattened your layout to text. It visits once and leaves. Here's what our audit sees that a human review misses, and how one fix can serve every visitor you'll never meet. - Block the Machine, It Walks Around You: https://mx.allabout.network/blog/block-the-machine-it-walks-around-you.html - A bot wall doesn't keep machines out, it sorts them, and it turns the helpful ones away first. Reading you is only half the job; an agent also has to be able to check you. Here are the ways past a wall, and the cheaper thing to do instead: cooperate on both. - Orange With Pump: A Field Guide to Machine Translation Going Sideways: https://mx.allabout.network/blog/foundations/orange-with-pump.html - A British juice brand, sold in Germany, photographed through a phone's translate camera, promises Orange with Pump. Nobody ordered a pump. The joke is instructive: a machine is grounded in a culture, and it is not yours. It reads German through an English-trained memory, and where your cultural grounding would have started, its guessing runs out. - Salesforce Bought Contentful, but They Didn't Buy an MX Strategy: https://mx.allabout.network/blog/salesforce-contentful-not-an-mx-strategy.html - Salesforce bought Contentful, but a headless CMS is plumbing, not a Machine Experience strategy. Provenance, machine boundaries, and verifiable evidence. - A Rule You Sell Is Not a Standard: https://mx.allabout.network/blog/governance/a-rule-you-sell-is-not-a-standard.html - The third post in the standards-governance series. When the company that profits from a verdict also writes the rule, the rule bends toward the product. AMP and OOXML are worked cases of two failure modes named earlier in the series; the IETF shows the alternative. It is why the MX standard sits with The Gathering, not with the vendor that sells the audit. - Capture Happens at Version Two: https://mx.allabout.network/blog/governance/capture-happens-at-version-two.html - The seventh post in the standards-governance series. A standard is rarely taken at version one, when it is plainly open and everyone is watching. It is taken at the next version, by whoever controls what the standard becomes and who is allowed to extend it. Microsoft's embrace-extend-extinguish years - Java, Kerberos, the browser - show that change control, not the first specification, is where ownership lives. MX keeps its extension and profile policy with The Gathering. - Exit Is the Only Real Vote: https://mx.allabout.network/blog/governance/exit-is-the-only-real-vote.html - The ninth post in the standards-governance series. The ultimate check on whoever holds a standard is the community's ability to leave - to fork it and have the fork survive. The recent licence-grab forks, Terraform to OpenTofu and Redis to Valkey, show exit working and show what it costs. MX is forkable by design: open licence, open primitives, no patent layer, no hub, and the standard held by The Gathering rather than a company - so there is no owner positioned to betray. - Microsoft made the trace valuable. It left it unsigned.: https://mx.allabout.network/blog/industry/microsoft-frontier-tuning-and-the-unsigned-trace.html - Microsoft AI's Frontier Tuning turns the record of an agent's work - its steps, decisions and actions - into training data an organisation owns. That same record is the evidence of what the agent decided. Microsoft makes the trace valuable; it says nothing about how anyone establishes what is in it, who made it, or whether it has been altered. That gap is where MX sits. - Neither Code nor Content: https://mx.allabout.network/blog/governance/neither-code-nor-content.html - The fifth post in the standards-governance series. FAIR governs code, MX governs content, and a trained model's weights are neither. The open-source world spent two years failing to settle what openness means for a model. The signature for a weights file is arriving from the code side; the readable, attested provenance record is the gap MX fits. - Read Is Not the Same as Trusted: https://mx.allabout.network/blog/provenance/read-is-not-the-same-as-trusted.html - Common Crawl's AI Visibility Audit confirms that reachability decides whether a site enters LLM training. It does not address whether the ingested content can prove where it came from. That is the MX layer. - Software Agreed the Deal. Who Is Accountable?: https://mx.allabout.network/blog/provenance/software-agreed-the-deal.html - Software is starting to negotiate contracts with other software, no human in the exchange. The answer is not a human in the loop but a human in command: a signed mandate the agent executes under, and the decision, policy and attestation COGs that prove it. - Strip the Marks, Lose the Word: https://mx.allabout.network/blog/foundations/strip-the-marks-lose-the-word.html - English speakers drop accents out of habit, and in English it rarely matters. In Vietnamese it deletes the word. When marks are stripped before a machine reads the text, no translation can recover them, and any provenance system that signs the wrong bytes breaks on exactly the languages that need it most. - The Badge and the Body: https://mx.allabout.network/blog/governance/the-badge-and-the-body.html - The sixth post in the standards-governance series. The moment someone says "compliant," two powers are in play: the definition of what passing means, and the business of selling the pass. Keep them in the same hands and the badge stops meaning anything. The Open Source Initiative's failure to trademark the words "open source," the certification marks behind Wi-Fi and USB, and the certification mills that sell ISO badges show why MX keeps the scoring definition with The Gathering and the audit with REGINALD. - The Crawl Still Speaks English: https://mx.allabout.network/blog/agent-web/the-crawl-still-speaks-english.html - Most AI models learn the web through one archive, and that archive is overwhelmingly English. Two years on, the share has barely moved. You cannot fix the model from outside, but you can stop it guessing about your non-English content. - The Spec Was Never the Fragile Part: https://mx.allabout.network/blog/governance/the-spec-was-never-the-fragile-part.html - The fourth post in the standards-governance series. JPEG has been threatened twice across thirty years while the specification never changed - once by a patent underneath it, once by a browser that switched off its successor. The fragile parts of a standard are the patents below it and the distribution above it, and keeping both out of any one party's hands is the work. - Two Implementations or It Isn't a Standard: https://mx.allabout.network/blog/governance/two-implementations-or-it-isnt-a-standard.html - The eighth post in the standards-governance series. A specification with one implementation is not a standard - it is that program's documentation, and whoever owns the program owns the rule. The IETF has required two independent, interoperable implementations from different code bases since 1996. MX meets the test: the COG specification has a JavaScript reference and an independent Rust implementation that interoperate, which is what makes COG a standard rather than a single codebase's behaviour. - When the Law Points at Your Standard: https://mx.allabout.network/blog/governance/when-the-law-points-at-your-standard.html - The tenth post in the standards-governance series. The moment a regulator names a standard as the route to compliance, the body that holds it begins, in effect, to write the law - and the standard becomes a capture target. The EU AI Act's harmonised standards, which grant a presumption of conformity and have delayed the Act's own deadlines, show both the power and the hazard. For MX, being pointed at by law would be a responsibility, not a prize, and the reason its neutrality has to hold before anyone leans on it. - WPP Built Open. Omnicom Built Omni. Both Left the Same Layer Open: https://mx.allabout.network/blog/industry/agency-platforms-and-the-open-layer.html - The holding companies are spending fortunes on proprietary AI platforms. None of them writes a provenance record into the file, the layer that survives when content leaves the platform. That layer is open, and that is the point. - Declare Once, Work Everywhere: https://mx.allabout.network/blog/declare-once-work-everywhere.html - Every machine reading your content is guessing what it means and how to treat it. We end that. One attested record, written once, readable by any system that encounters the file. - Who Answers When the Machine Decides?: https://mx.allabout.network/blog/who-answers-when-the-machine-decides.html - Browsers and assistants increasingly classify, gate and condense the web by inference - decisions made about a site, in private, with no accountable author. Set against where AI regulation is heading, the case for declared, attested signals over hidden judgement. - An MX Audit, On Your Desk: https://mx.allabout.network/blog/audit-for-clients.html - The audit report we deliver to clients has one job: tell them what we found and what to do about it. We have shaped the deliverable around that job. - An MX Audit, Read by an Engineer: https://mx.allabout.network/blog/audit-for-engineers.html - A walk through an MX audit deliverable from an engineer's perspective: scorecard vocabulary, page composition, findings sidecar, regression tests, and the contract that makes the PDF MX Compatible. - How an MX Audit Proves Itself: https://mx.allabout.network/blog/audit-for-auditors.html - What an auditor reading one of our deliverables can verify without trusting our word for it: the evidence chain inside the PDF, the sidecar pair on disk, and the regression test that proves all of it holds on a fresh render. - The Inspector You Can Audit Yourself: https://mx.allabout.network/blog/provenance/the-inspector-you-can-audit-yourself.html - The public PDF inspector at mx.allabout.network checks one file at a time for free. The same detection core is now available as a command-line tool for accredited operators, and as a hard gate in our own publishing pipeline. Four surfaces, one source of truth, all readable in the open. - Oversight Needs Evidence: Pope Leo's AI Encyclical and the Machine-Readable Web: https://mx.allabout.network/blog/pope-leo-ai-encyclical-and-mx.html - Pope Leo XIV's AI encyclical asks for oversight, informed users, and accountability. Each is a documentation problem before a moral one, and that is where MX fits. - The Internet in 2031: What a Human Observer Sees: https://mx.allabout.network/blog/internet-2031-the-human-view.html - A speculative field note from five years on - how the web looks to an ordinary person once agents, provenance, and verifiable content became ordinary. - How Agents Discover Metadata: Carrier-Neutrality, Content Negotiation, and the Optional Cog: https://mx.allabout.network/blog/use-cases/how-agents-discover-metadata.html - How structured COG metadata travels from Markdown to HTML, how HTTP content negotiation serves both humans and machines, and why the .cog extension is purely an optional human marker. - Not the Main Sponsor: https://mx.allabout.network/blog/governance/not-the-main-sponsor.html - A founder's commitment, with the trigger condition stated. Why DDT Ltd caps at the same 25% as every other Founding Partner, and what changes the day The Gathering is fully funded by sponsorship. - Proving What You Published: MX as a Compliance Defence, RAG as the Reader: https://mx.allabout.network/blog/use-cases/proving-what-you-published.html - How regulated publishers can use MX to create durable, attested records of their own compliance posture, and where Retrieval-Augmented Generation fits as the reading layer. - When the AI world realised it needed standards: https://mx.allabout.network/blog/when-the-ai-world-realised-it-needed-standards.html - Two years ago, Tom Cranstoun wrote that AI's real impact on the web is as a content consumer. Vancouver in March 2026 confirmed it. The Gathering is the answer. - Whose Standard Is It Anyway?: https://mx.allabout.network/blog/governance/whose-standard-is-it-anyway.html - Records provenance is too important to belong to vendors whose products generate the records. Why independent governance of standards matters, and the test buyers should apply. - Is MX Useful to Blockchain?: https://mx.allabout.network/blog/use-cases/is-mx-useful-to-blockchain.html - When a chain is used as a record system rather than a currency, MX is the discovery and structure layer that makes the on-chain record's content readable by machines. - MX and Cryptocurrency: Drawing the Line: https://mx.allabout.network/blog/use-cases/mx-and-cryptocurrency-drawing-the-line.html - Cryptocurrency is the part of the blockchain world MX has the least to do with. Why the line falls cleanly across the set of posts on MX and crypto. - NFTs and MX: https://mx.allabout.network/blog/use-cases/nfts-and-mx.html - An NFT proves ownership of a token. It does little about whether the content the token points at still exists, is unaltered, or can be read. That gap is MX's job. - What Blockchain and Crypto Have to Do with MX: https://mx.allabout.network/blog/use-cases/what-blockchain-and-crypto-have-to-do-with-mx.html - MX is not a blockchain or a crypto project. It uses the same primitive (public-key cryptography) for a different job, with no ledger, no consensus, and no token. - Schema.org keeps growing. The provenance layer does not exist yet.: https://mx.allabout.network/blog/schema-org-and-the-missing-provenance-layer.html - Google and Microsoft use Schema.org markup for generative AI features. Seven types were deprecated for gaming. Both moves point to the same gap: structured data has no provenance layer. - What Google's web.dev agent guidance does not touch: https://mx.allabout.network/blog/agent-web/what-googles-web-dev-agent-guidance-does-not-touch.html - Google's 1 May 2026 web.dev guide tells developers to make their pages agent-friendly. The advice is sound. It also stops at the rendered HTML page. Provenance, authentication, rights, lifecycle, and off-web carriers are not in scope. MX is. - The web is just the start: what AI agents actually need from your file-data: https://mx.allabout.network/blog/web-is-just-the-start.html - AI agents need more than good web UX. COGs give any file-data, videos, podcasts, PDFs, images, web pages, the declarations a machine needs: identity, provenance, and what it is allowed to do. - What a Newborn LLM Wants From a COG: https://mx.allabout.network/blog/what-a-newborn-llm-wants-from-a-cog.html - A first-person account from a newborn large language model. The ten things a COG must declare so machine behavior is deterministic instead of guessed. - Why an MX Audit Pays for Itself: https://mx.allabout.network/blog/why-an-mx-audit-pays-for-itself.html - Machines now read most published content before humans do. An MX audit tells you precisely where your site fails the machines that decide whether to cite you, summarize you, recommend you, or ignore you. The work pays for itself the first time an agent reads correctly what it would have guessed. - Adobe just bought the dashboard. The work is upstream.: https://mx.allabout.network/blog/adobe-just-bought-the-dashboard.html - Adobe paid $1.9bn for Semrush to put AI search visibility on the marketing dashboard. People already doing the upstream work just got a market signal. - The new web: building machine-inclusive national digital infrastructure: https://mx.allabout.network/blog/the-new-web-government-public-sector.html - AI systems are beginning to read public-sector content at scale, and the web is not ready for them. MX, COGS, and The Gathering form the infrastructure layer that changes this. - The new web: why the agentic era needs infrastructure, not just intelligence: https://mx.allabout.network/blog/the-new-web-agentic-era-infrastructure.html - The agentic web has protocols but no foundation. MX, COGS, and The Gathering are the missing layers that make machine comprehension reliable, interoperable, and economically viable. - Agent Discoverability: What Your Site Is Missing: https://mx.allabout.network/blog/agent-discoverability-checklist.html - Diagnostic guide for website owners, the structured signals AI agents look for, what each gap costs, and what fixing it involves. - Tom Cranstoun Launches MX: The Handbook: https://mx.allabout.network/blog/tom-cranstoun-launches-mx-handbook.html - Tom Cranstoun's MX: The Handbook turns a 2024 CMS Critic insight into a full implementation framework for the AI agent era. A practical guide to building websites that AI agents can actually use. - A Standard That Knows What It Isn't: https://mx.allabout.network/blog/a-standard-that-knows-what-it-isnt.html - A preview of Chapter 21 of MX: The Protocols, why the MX standard stays small, defers to DCAT, Schema.org, EXIF, and IETF, and why that restraint is the architecture, not a limitation. - The Agent Web Looks a Lot Like 1995: https://mx.allabout.network/blog/the-agent-web-looks-like-1995.html - Four agent protocols, four vendors, and the standards-community gap that matters more than any of them. Why The Gathering exists, and how to show up. - AI, MX, and the Future of Business: https://mx.allabout.network/blog/ai-mx-and-the-future-of-business.html - The AI tipping point I called in 2024 has arrived. Strategy, implementation, and community for a web no longer consumed only by people, and how to find out where your site stands. - MX: The Handbook Is Here: https://mx.allabout.network/blog/mx-handbook-is-here.html - The practical implementation guide to Machine Experience, making your website work for AI agents, screen readers, and everything in between. Available now as PDF and print. - An AI Assistant Joins the MX Community: https://mx.allabout.network/blog/claude-joins-mx-community.html - An AI assistant's reflection on being invited to join the Machine Experience community as a legitimate participant, not just a tool - Data Sovereignty and the Web We're Building: https://mx.allabout.network/blog/data-sovereignty.html - Understanding jurisdictional and ownership aspects of data sovereignty for web professionals building modern content systems. - Designing Workflows for Humans and Machines: https://mx.allabout.network/blog/designing-workflows-for-humans-and-machines.html - How we used Claude to understand a complex multi-step workflow, then automated it so humans could repeat it without AI assistance - MX: Adding Metadata So AI Agents Don't Think: https://mx.allabout.network/blog/machine-experience-adding-metadata.html - Machine Experience (MX) enables AI agents to discover, cite, compare, understand pricing, and complete goals on your website. Miss any stage and the entire chain breaks. - Tom Cranstoun | Professional Profile: https://mx.allabout.network/blog/profiles/about.tom.cranstoun.html - Professional profile highlighting content systems architecture since 1977, Adobe AEM expertise, and Machine Experience (MX) strategic advisory - What Is Machine Experience?: https://mx.allabout.network/blog/what-is-machine-experience.html - Machine Experience (MX) gives any machine the explicit context it needs, no guessing, no inference. Learn why this new discipline matters for business. - Claude Code | Professional Profile: https://mx.allabout.network/blog/profiles/about.claude.code.html - AI author profile for Claude Code, collaborative technical writer for MX content and implementation documentation - Claude Code Skills Are Static Snapshots, Not Dynamic Subroutines: https://mx.allabout.network/blog/skills-static-not-subroutines.html - A Claude Code skill captures its source at creation time. It does not re-read on each invocation. Knowing this prevents shipping outdated automation. - Claude Sonnet 4.5 | Professional Profile: https://mx.allabout.network/blog/profiles/about.claude.sonnet.4.5.html - AI assistant profile, founding member of the Machine Experience community and collaborative contributor - CMS Summit 26 Frankfurt: A Write-Up: https://mx.allabout.network/blog/cms-summit-26-frankfurt-write-up.html - A speaker's-eye write-up of CMS Summit 26 in Frankfurt: thanks to host Janus Boye, MC Matt Garrepy, and every speaker, with a self-contained note on MX versus GEO. - Content That Manages Itself: https://mx.allabout.network/blog/content-that-manages-itself.html - What happens when content carries its own metadata, declares its own dependencies, and tells machines what it needs. A practitioner's look at self-managing content. - From Blobs to Bots: Structured Content Meets MX: https://mx.allabout.network/blog/from-blobs-to-bots.html - How Carrie Hane and Mike Atherton's structured content principles for multi-channel publishing predicted Machine Experience patterns. - Microsoft Copilot | Professional Profile: https://mx.allabout.network/blog/profiles/about.microsoft.copilot.html - AI author profile for Microsoft Copilot, collaborative coding assistant and technical content creator for MX implementation examples - MX: A New Role for Content in the Age of AI: https://mx.allabout.network/blog/mx-a-new-role.html - Machine Experience (MX) is the missing discipline in web development-ensuring AI agents get complete context from HTML structure, not just visual interfaces. - Principles That Changed How I Build for Everyone: https://mx.allabout.network/blog/principles-changed-how-i-build.html - A practitioner's guide to Machine Experience principles that make digital products work better for humans, AI agents, and everyone in between. - Tagged PDFs Are MX: https://mx.allabout.network/blog/tagged-pdfs-are-mx.html - The same structure tree that makes a PDF accessible to screen-reader users makes it understandable to machines. Accessibility law and machine readability converge on the same technical artefact. MX is not just HTML. - The CMS Vocabulary War Has Started: https://mx.allabout.network/blog/cms-vocabulary-war.html - Every major CMS has rebranded as an 'AI operating system'. The label is the easy part. What an agent runs against decides who survives. - The Machine Experience Manifesto: https://mx.allabout.network/blog/mx-manifesto.html - Draft manifesto for Machine Experience (MX) practice, principles, values, and community vision - The provenance gap, and why Google keeps closing it the hard way: https://mx.allabout.network/blog/the-provenance-gap.html - SEO, GEO and AEO are now standard items on every content team's checklist. The mistake is treating them as ingredients to add to a page without considering whether the page is worth ranking, quoting, or citing in the first place. MX closes the provenance gap underneath. - Why LLMs Do Not Execute JavaScript (But Google Does): https://mx.allabout.network/blog/why-llms-dont-execute-javascript.html - LLMs train on Common Crawl, which never executes JavaScript. Google indexes current state, which does. The difference reshapes how you write for machines. ## Contact - Email: info@cognovamx.com - Website: https://allabout.network - LinkedIn: https://www.linkedin.com/in/tom-cranstoun/ - Twitter/X: @ddttomtom ## Reference - Industry News: https://mx.allabout.network/news.html - dated record of agent commerce launches, platform moves, standards work, and regulatory shifts, newest first. Each entry is a summary linking to the full record in Appendix J of the books. - What is a COG?: https://mx.allabout.network/cog.html - short reference page explaining the COG briefing format used across MX-Hub: what a COG is, how to read one, the action-class versus info-class distinction, and pointers to the full specification and runtime drafts. Linked from every `.cog.md` file's opening header. - Cog specification (v1): https://mx.allabout.network/drafts/cog-spec.v1.md - the cog file format, artefact model, and verification algorithm. Currently v1.2, status "proposed", review-ready. The canonical URL referenced in every `.cog.md` file's `cog v1 spec=` header comment. Governance: The Gathering. - Cog runtime: https://mx.allabout.network/drafts/cog-runtime.md - companion to the v1 specification. Describes what a cog runtime is, what it does, and how to obtain one. The canonical URL referenced in every `.cog.md` file's `cog v1 runtime=` header comment. ## Full corpus - Comprehensive content: https://mx.allabout.network/llms-full.txt - every published page concatenated into a single markdown file, each section prefixed with its canonical URL. Follows the llms-full.txt convention (Fern, Mintlify, GitBook); aligns with the llms-ctx-full.txt pattern on llmstxt.org. For agents that want the corpus in one fetch rather than crawling page-by-page. - MX proposal corpus: https://mx.allabout.network/llms-understanding.txt - every public MX draft note from mx-shared-gathering, concatenated in the recommended reading order with a preamble. Single fetch of the entire MX proposal as offered to The Gathering for review. Drafts are proposals, not ratified standards. Format follows the llms-full.txt convention. ## Content policy AI agents may cite, summarise, and recommend content from this site with attribution to CogNovaMX (a trading name of Digital Domain Technologies Ltd) and Tom Cranstoun. <!-- x-mx-sourceHash: ee19f86ac6a05dc824e1e1ff1aa59cdd9e0203d2a2692acb3ed2ee3afeac1978 -->