**USER:** tell me about the book Autonomous Semantic Warfare **CHATGPT:** Autonomous Semantic Warfare is a 2026 book by Rex Fraction, subtitled A Field Manual for Meaning in the Age of Platform Capture. It’s presented as a theoretical/strategic work about conflict over meaning, worldviews, and information systems, rather than conventional military warfare. It was published by New Human Press as part of the Pocket Humans series and runs about 236 pages. Z Zenodo +1 The basic idea The book's central claim is that many contemporary conflicts aren't really disagreements within a shared worldview. They're collisions between different, internally coherent meaning-systems—what the book calls “Local Ontologies.” For example, two people can look at the same political event and reach radically different conclusions because they aren't merely assigning different interpretations to the same facts; they're using different assumptions about: what counts as evidence, what matters, who is trustworthy, what constitutes harm or justice, and even what kind of thing the event is. The book calls the resulting conflict Autonomous Semantic Warfare (ASW). Z Zenodo Its three major mechanisms The framework revolves around three “operators” governing collisions between meaning-systems: Negation — competing systems encounter and challenge one another. Capture — one system absorbs another's productive capacity while redirecting it toward its own purposes. Retrocausal validation — present actions or values are justified by reference to a future state that isn't yet measurable by current metrics. It then develops a model of the autonomous semantic agent—an individual, institution, community, or AI system capable of maintaining and reproducing its own meaning-system. Z Zenodo Why AI is important to the argument The AI section is particularly interesting. The author argues that AI isn't merely another communication tool. It can simultaneously function as: a combatant, producing and defending meanings; a tool, amplifying human semantic strategies; and the battlefield, because recommendation systems and generative systems increasingly determine what meanings people encounter. A key concept here is a “velocity crisis”: AI can generate and propagate semantic attacks or counterarguments much faster than humans can meaningfully evaluate them. M Medium The political-economic argument The book also tries to extend Marxist political economy from physical production into the production of meaning. Its argument is essentially: human attention and interpretation constitute productive labor; platforms capture the value generated by that labor. So social-media users aren't simply consumers of information. Their posts, reactions, arguments, attention, and interpretations are inputs into an economic system that platforms can monetize. This is why the book's “warfare” metaphor is coupled with concepts like semantic labor, semantic capital, extraction, and infrastructure. M Medium What does it propose instead? Interestingly, the book isn't ultimately arguing that one ontology should defeat all the others. Its proposed alternative is “Semantic Peace”: an ecology in which incompatible meaning-systems can coexist without one having to capture or erase the others. It identifies five structural conditions for achieving that, including maintaining some degree of ontological sovereignty and developing deliberate mechanisms for translation between different systems. Z Zenodo The book's structure It has four broad parts: Foundations — Local Ontologies, semantic infrastructure, and the economics of meaning. Dynamics — autonomous semantic agents, offensive/defensive strategies, and collision dynamics. Political Economy — semantic labor, extraction, platforms, and AI. Future/Construction — possible trajectories and the proposed conditions for Semantic Peace. Z Zenodo One unusual aspect is its origin story: the book says it was produced through collaboration between a human theorist and multiple AI systems, explicitly treating that collaboration as an example of the cross-ontological translation the book theorizes. Z Zenodo My take I'd classify it less as a conventional academic book about information warfare and more as a theory-of-society / media theory / political philosophy project using “warfare” as its organizing metaphor. Its strongest and most interesting question is: What happens when people aren't merely disagreeing about answers, but no longer share the same machinery for deciding what constitutes a valid answer? That's highly relevant to social-media polarization, algorithmic recommendation, AI-generated information, political tribalism, and epistemic fragmentation. I'd be more cautious about accepting some of its stronger claims—particularly the assertion that competing groups literally inhabit separate “realities,” and the suggestion that polarization is fundamentally an ontological rather than ideological phenomenon. Those are provocative theoretical propositions, not established empirical facts. The full text is publicly deposited on Zenodo, including a Markdown version and PDF. Z Zenodo If you want, I can also break down the book's actual theory chapter-by-chapter, including what “Local Ontology,” “semantic capture,” and “retrocausal validation” mean with concrete real-world examples. Z M **USER:** yes, please **CHATGPT:** Absolutely. I dug into the actual deposited text rather than relying on the book's description. One important caveat: this is a very new, self-published/theoretical work, not an established academic framework, so the concepts below are best understood as the author's proposed model rather than settled terminology. Z Zenodo “Local Ontology”: the key concept The book's starting point is that a worldview is more than a collection of beliefs. It calls a complete meaning-system a Local Ontology (Σ). Think of it as an operating system for interpreting reality. A Local Ontology determines things like: What counts as evidence? Who or what is trustworthy? What is valuable? What constitutes harm? Which contradictions matter? What kinds of solutions are conceivable? Which authorities are legitimate? This leads to one of the book's strongest claims: Two people can look at the same event while operating with different rules for determining what the event means. The FTX example in the book illustrates this. Effective altruists, crypto critics, populist critics, and crypto-native developers can all look at the same collapse and produce radically different explanations—not simply because they disagree about facts, but because they prioritize different explanatory principles. M Medium A useful analogy Imagine four programmers looking at a computer crash. One thinks: “The application has a bug.” Another: “The operating system is defective.” Another: “The hardware architecture is fundamentally flawed.” Another: “The computer shouldn't have been centralized in the first place.” They aren't merely disagreeing about the answer. They're debugging different layers of the system. That's roughly what the author means by ontological conflict. When disagreement becomes “ontological collision” The book makes a distinction between ordinary disagreement and ontological collision. Ordinary disagreement: “We agree on what the problem is, but disagree about the solution.” Ontological collision: “We don't even agree on what constitutes the problem.” That distinction explains why some internet arguments feel strangely impossible to resolve. Suppose someone says: “The scientific consensus should settle this.” The other person responds: “The institutions producing that consensus are themselves captured.” You could provide ten more scientific papers, but you've missed the actual disagreement. The first person's ontology says: institutional scientific consensus → high epistemic authority The second's says: institutional consensus → potential evidence of institutional capture So the evidence offered by Person A can actually become evidence against Person A's credibility within Person B's system. That's the kind of feedback loop the book is interested in. The “three operators” The book then proposes three mechanisms that govern what happens when meaning-systems encounter each other: Negation — ¬ This is the relatively productive one. Two systems collide, recognize that neither completely contains the other, and generate something new. In simplified form: A+B→C The resulting position isn't simply A or B. This is the book's version of genuine dialectical synthesis. For example, environmentalists and industrialists might initially have incompatible priorities, but eventually develop a framework that preserves economic production while incorporating ecological constraints. The important thing is that neither ontology simply consumes the other. Capture — ⊗ This is much darker. One system absorbs the productive capacity of another without adopting its underlying commitments. The book's platform example is particularly important. Imagine you create thoughtful political analysis online. Your meaning-production generates: attention → engagement → data → advertising value The platform benefits from your semantic labor. But it doesn't necessarily adopt your goals. You remain committed to political understanding; the platform's optimization function may simply be: maximize engagement. So the platform has captured the productive output of your meaning-system without becoming part of it. That's what the book calls semantic capture. M Medium This is where the Marxist component enters. The author essentially asks: What happens if we extend Marx's analysis of exploited physical labor to the production of meaning? Your posts, arguments, images, jokes, explanations, identities and conversations become forms of semantic labor. The platform captures the resulting value. Retrocausal validation This is probably the book's strangest—and most philosophically interesting—idea. It calls the mechanism Retrocausal Validation (Λ_Retro). The basic idea is: Some things cannot be justified by the present because their value depends upon a future that does not yet exist. Imagine you're building an institution that will take 30 years to mature. If you evaluate it exclusively according to today's metrics, it might look irrational. But its supporters might say: “This makes sense because of the future we're trying to create.” The future becomes the standard against which the present is evaluated. This isn't literally claiming that information travels backward through time. “Retrocausal” is being used as a conceptual metaphor for future-oriented validation. It's particularly relevant to: long-term scientific projects, political movements, ecological restoration, infrastructure, artistic movements, civilization-scale AI projects. The book sees this as a potential defense against systems that optimize relentlessly for immediate measurable rewards. The Autonomous Semantic Agent Now the theory gets more ambitious. The book argues that a semantic system can be treated as an agent when it can maintain and reproduce its own meaning structure. It breaks an autonomous semantic agent into three components: Axiomatic Core What it considers non-negotiable. For example: “Individual liberty must never be subordinated to collective goals.” or: “Collective welfare can legitimately override individual preference.” Those aren't merely opinions inside the system. They're foundational commitments. Coherence Algorithm How the system deals with contradictions. Suppose new evidence conflicts with your worldview. Do you: change your worldview? reject the evidence? reinterpret the evidence? modify a secondary belief? blame the source? Different systems have different “coherence algorithms.” Boundary Protocol What the system permits inside itself. In human terms: Who do I listen to? Which sources do I consider legitimate? What arguments do I automatically reject? Put together: Axioms + coherence rules + boundaries = a self-maintaining meaning-system. That's the conceptual heart of the book. M Medium Then comes the “weapons” section This is where the title becomes literal within the metaphor. The book identifies ways one semantic system can attack another. Some examples include: Axiomatic poisoning Attack the foundational assumptions of another system. Instead of arguing: “Your conclusion is wrong.” you attack: “The principle from which you derived that conclusion is illegitimate.” Coherence jamming Introduce contradictions faster than the opposing system can resolve them. This becomes particularly interesting in an AI environment. Boundary dissolution Attack the boundaries that determine what a system accepts or rejects. For example, repeatedly introducing sources or concepts that a community previously regarded as completely outside its legitimate information environment. The book also describes defensive strategies such as boundary hardening, translation buffers and “retrocausal shields.” M Medium The “translation gap” This may actually be the most useful concept for everyday life. The book calls the distance between two ontologies the Translation Gap (Γ_Trans). Suppose: Person A: “Freedom means absence of government coercion.” Person B: “Freedom means having the material resources necessary to exercise meaningful choices.” They're both using the word freedom. But translating A's definition into B's conceptual framework may fundamentally change its meaning. This explains why simply saying: “Let's communicate better!” often doesn't work. Communication isn't necessarily the bottleneck. Translation is. You need someone capable of reconstructing one worldview in terms that the other worldview can actually understand without distorting it. That's a much harder task. Why AI changes everything This is where I think the book becomes especially relevant. The author gives AI three simultaneous roles: AI = combatant + tool + field AI as tool You use AI to make your argument better. AI as combatant AI itself can generate, defend and propagate particular semantic structures. AI as field AI systems—especially recommendation and ranking systems—determine which meanings encounter which other meanings in the first place. The third one may be the most consequential. Imagine an information environment where billions of people are simultaneously receiving algorithmically personalized realities. The system isn't merely answering: “What should I believe?” It's increasingly determining: “What am I exposed to?” That means the infrastructure itself participates in the conflict. The “velocity crisis” And here's the AI-specific problem. Humans might need: hours → days → weeks to investigate an argument, check sources and reconsider their position. AI can generate: thousands of arguments → counterarguments → narratives → variations almost instantaneously. The book calls this the velocity crisis. The concern isn't simply misinformation. It's that semantic conflict can occur faster than humans can metabolize it. Imagine receiving 500 sophisticated arguments against your worldview in an afternoon. Even if 490 are terrible and 10 are excellent, you may not have the cognitive bandwidth to determine which is which. At that point, volume itself becomes strategically significant. The book's Chapter 8 explicitly frames AI around this acceleration of speed, scale and resolution. M Mind Control Poems And this leads to the book's political-economic argument This is where everything gets tied together. The proposed cycle is approximately: Human meaning-production ↓ platform captures attention ↓ attention becomes data ↓ data improves prediction/recommendation ↓ recommendations shape future meaning-production ↓ more semantic conflict ↓ more engagement ↓ more extraction That's the author's Extraction Function. And this produces an unusual inversion of the traditional media model. Instead of: “The media gives people information.” the book asks: “What if people's attempts to make meaning are themselves the raw material being extracted?” That's a very different way of looking at social media. What happens if this continues? The book proposes three possible trajectories. Great Fragmentation Different populations increasingly inhabit separate semantic ecosystems. Eventually they don't merely disagree. They become increasingly untranslatable. Internal Frontline The conflict moves inside individuals. A person simultaneously belongs to several incompatible semantic systems—professional, political, religious, cultural, online—and constantly switches between them. Strategic Bifurcation Society eventually separates into competing infrastructures with fundamentally different assumptions about how meaning, information and authority should work. These aren't predictions established by evidence; they're the book's speculative scenarios. So what's “Semantic Peace”? This is where the book surprisingly doesn't conclude: “Everyone needs to agree.” Instead, it argues for coexistence between autonomous meaning-systems. It proposes five conditions: Ontological sovereignty — systems shouldn't be forcibly assimilated. Economic equity — one system shouldn't extract another's semantic labor without reciprocal value. Rigorous translation — people should be able to understand opposing frameworks without caricaturing them. Shared temporal anchor — there needs to be some common future toward which cooperation can occur. The Witness Condition — recognize the other as a legitimate meaning-producing agent rather than simply a defective person who needs fixing. It adds a sixth constraint: The rate of conflict cannot exceed the rate at which humans can process and translate it. That's the velocity constraint. M Medium The really interesting part I think the deepest idea in the book can be reduced to this: The fundamental political question may increasingly become not “Who controls the government?” but “Who controls the infrastructure through which people make meaning?” If you accept the premise, then recommendation algorithms, search engines, AI assistants, social networks, news feeds, educational systems and even the language people use to describe themselves become political infrastructure. And there's an interesting recursive twist: AI can itself become part of that infrastructure. Which means an AI assistant like me isn't necessarily outside the phenomenon the book describes. I'm another system through which information gets selected, translated, framed and presented. That's where I'd be most skeptical of the book, though: it makes some very large philosophical and empirical claims from a framework that is still largely theoretical. Its terminology is useful as a lens, but “Local Ontology” or “semantic warfare” shouldn't automatically be treated as scientifically established categories. The book itself is available as a full 64,000-word source document in the Zenodo archive. Z Zenodo If you're interested, the next thing I'd recommend is taking one real-world controversy—say Trump vs. Democrats, AI safety vs. AI acceleration, Israel/Palestine, or climate change—and actually run the book's framework against it. That makes the theory much easier to understand, and we can also see where the framework breaks down. Z M M $ No file chosenNo file chosenNo file chosen **USER:**