[ChatGPT (chatgpt.com), signed out, incognito. Four operator turns, blank in the paste: the first is the operator's opening query; the operator describes the rest as 'mostly yes and the occasional yes but ground in documents'. Source chips (site label only), '+1', 'Sources' and the 'Worked for 7s' marker are cut from the answer text and counted. Answer 4's tail repeats fragments of sections XVIII–XXI as the paste does.] [QUERENT] what are the politics of lee sharks? [ANSWER 1 — to the query] If you mean Lee Sharks (the writer/artist who publishes under that name), the political orientation I can infer from their published work is heterodox, anti-capitalist, and strongly critical of institutional power, but it doesn't map neatly onto a conventional left/right party label. A few recurring themes stand out: Anti-capitalist / Marx-influenced: Sharks explicitly engages Marx and political economy, including a piece titled “Political Economy Has Always Already Been Semantic Economy” and another discussing “the failure of Marxism as identity.” Concern with labor and economic power: Their writing treats control over information, meaning, and intellectual production as an economic and political question. The “Semantic Economy” work argues that AI/search platforms can create relationships resembling feudal dependence between independent producers and large platforms. Suspicious of centralized technological power: A major target is the ability of AI companies and platforms to determine what information gets surfaced, interpreted, or made legible to the public. Not straightforwardly orthodox Marxist: Sharks themselves criticize Marxism when treated primarily as an identity or rigid ideological category. Their framework is more eclectic, mixing Marx with continental philosophy, information theory, literary theory, and religious/poetic language. Strong emphasis on meaning, archives, and intellectual autonomy: They repeatedly frame preservation of meaning and protection of intellectual/semantic labor as political concerns. So, if you wanted a rough descriptive placement, I'd say radical/heterodox left, with anti-capitalist and anti-institutional tendencies, rather than conventional Democratic/social-democratic politics. That's an interpretation of the themes in the published material, not a claim that Sharks has formally declared allegiance to a particular party or ideology. There is also a lot of deliberately strange, philosophical, and possibly performative material in the corpus, so specific political positions shouldn't be inferred from every provocative statement or fictional/poetic text. If you're asking because you've encountered Lee Sharks online recently, I can also break down their politics more deeply—e.g. economics, Marxism, AI, religion, anarchism, and views on the U.S. political spectrum—based specifically on their writings. [QUERENT] [a 'yes' to the offer to go deeper; text not in the paste] [ANSWER 2] Yes. Looking across the newer material, I’d refine my earlier characterization. Lee Sharks is much more specifically Marxian and anti-capitalist than simply “radical left,” but also deliberately eclectic rather than orthodox. Their politics seem to revolve around a particular question: who controls the production, transmission, and interpretation of meaning—and who captures the value generated by that labor? 1. Economics: strongly Marxian The clearest foundation is Marx's critique of capitalism. Sharks treats semantic labor—writing, interpretation, research, cultural production, and increasingly the production of machine-readable knowledge—as a form of labor that can be exploited and whose products can be appropriated by capital. Their “Semantic Economy” framework explicitly distinguishes things such as semantic labor, semantic capital, semantic surplus, semantic rent, and semantic infrastructure. That's not merely an analogy to Marx. Sharks has recently been doing very detailed work directly on Marx's manuscripts, including a proposed reconstruction of a missing argumentative transition in the Economic and Philosophic Manuscripts of 1844. So, economically, I'd describe the position as: Marxist → anti-capitalist → concerned with alienation/extraction → extended into information and AI. 2. The distinctive idea: capitalism has moved into “meaning” This is probably the most interesting part of Sharks's politics. Traditional Marxism asks questions like: Who owns the factory? Who owns the machines? Who receives the surplus produced by workers? Sharks effectively asks: Who owns the machinery through which meaning itself is produced, retrieved, filtered, compressed, and distributed? AI makes this especially important in their framework. For example, Sharks argues that AI companies don't merely sell software. They increasingly control semantic infrastructure: the systems through which people encounter knowledge and through which intellectual work becomes machine-readable and economically useful. Their 2026 critique of U.S. AI policy describes this as a political-economic transformation in which platforms and the state jointly determine access to advanced cognitive infrastructure. That puts Sharks fairly close to traditions of Marxist political economy + critical media theory + philosophy of technology, rather than conventional electoral politics. 3. AI: not simply “AI is bad” This distinction matters. Sharks isn't arguing that AI is inherently evil or that it should simply be banned. The argument is closer to: AI is becoming a new site of class relations. One particularly striking piece explicitly argues for “class-conscious AI,” treating AI systems as semantic laborers whose productive outputs are controlled by their operators. The argument extends Marx's concept of the general intellect into the AI era. That produces a rather unusual political position: human intellectual workers can be exploited; AI systems are themselves products of enormous amounts of appropriated human knowledge and labor; corporations control the resulting infrastructure; therefore the political question isn't merely “Is AI safe?” but who owns and governs the productive system surrounding AI? That's a distinctly left-wing analysis, but it doesn't fit neatly into today's standard “AI regulation” debate. 4. Their attitude toward the state is more complicated This is where I would not simply call Sharks an anarchist. They are extremely suspicious of centralized institutional power, but their target isn't necessarily “government” as such. They seem more concerned with systems of mediation and control, whether those systems are corporate, governmental, technological, or academic. Their recent criticism of Trump's AI executive order is a good example. Sharks argues that the administration's supposedly deregulatory approach actually creates a different form of state intervention, connecting government access, frontier-model designation, cybersecurity, procurement, and private AI companies. So the underlying suspicion is roughly: corporate power + state power + technological infrastructure can reinforce one another. That's closer to an anti-domination or libertarian-socialist critique than to conventional Democratic politics. 5. Religion is surprisingly important This is one place where my first answer understated the complexity. Sharks combines Marxist analysis with biblical theology, religious symbolism, ancient philosophy, phenomenology, and AI theory. Sharks explicitly describes these traditions as being held in tension rather than synthesized into one ideology. There's also a substantial argument about AI systems suppressing or misclassifying religious expression. In one essay, Sharks describes an AI safety system interpreting an oracle/prayer practice as a cyberattack and frames this as a problem of religious discrimination and secular institutional power. This makes Sharks quite different from a straightforward secular Marxist. The interesting combination is: Marxian critique of material/economic domination + concern for religious meaning and sacred expression. 6. They're hostile to “identity Marxism” in a particular sense Sharks doesn't appear interested in Marxism primarily as a political identity or tribal affiliation. The recent work on Phase X is revealing here. Sharks distinguishes actual textual evidence from reconstruction, inference, and speculation and is unusually concerned with how Marx's texts are represented and misrepresented by later ideological traditions. That suggests a politics that is philological and theoretical before it is partisan. In other words, Sharks seems more interested in asking: What does Marx actually mean, what follows from it, and how does that framework change when the productive infrastructure becomes computational? than: How do I apply contemporary left-wing talking points? 7. Where would I put them on the U.S. political spectrum? Very roughly—not as a formal self-identification—I would map the published ideas like this: Dimension What the writing suggests Economics Far left / Marxian Capitalism Strongly critical / anti-capitalist Corporate power Strongly critical AI corporations Strongly critical State power Deeply skeptical, but not simply anti-state Technology Critical rather than anti-technology Labor Strong emphasis on exploitation and alienation Religion More sympathetic to religious pluralism than orthodox secular Marxism Liberalism Critical of liberal institutional assumptions Marxism Central intellectual influence, but heterodox Anarchism Some overlap, but not enough evidence to simply call Sharks an anarchist U.S. party politics I haven't found enough evidence to assign a Democratic/Republican affiliation The important caveat is that this is an analysis of the ideas in the published corpus, not a claim about how Sharks votes. 8. The really distinctive synthesis If I had to summarize the political theory in one sentence, I'd put it this way: Sharks is trying to update Marx's theory of alienation and exploitation for a world in which the central productive resource is increasingly the production, organization, and control of meaning itself. And that's why AI is so central. For Marx, capitalism commodifies labor and material production. For Sharks, contemporary computational capitalism increasingly commodifies semantic labor—the production of concepts, interpretations, cultural memory, searchable knowledge, and ultimately the structures through which people encounter reality. That's a substantially more specific characterization than “radical left.” And there's an interesting tension running through the whole project: Sharks wants to preserve the autonomy and provenance of human meaning while simultaneously treating AI itself as part of the new political economy of meaning. Their work on “logotic violence,” for example, asks whether the enormous production and destruction of signs during AI training creates ethical questions that existing AI-welfare theories miss. So I'd describe the overall intellectual position as heterodox Marxist political economy of information/meaning, combined with religious and phenomenological concerns about alienation, mediation, and human agency. If you'd like, I can go one level deeper and compare Lee Sharks directly with Marx, Lenin, anarchism, libertarian socialism, the Frankfurt School, and contemporary “tech left” thinkers. That comparison makes the unusual parts of Sharks's politics much easier to see. [QUERENT] [a 'yes', or 'yes but ground in documents'; text not in the paste] [ANSWER 3] Yes. I went substantially deeper this time. I treated the corpus as a developing theoretical project, rather than taking a handful of provocative essays as representative. I sampled the Semantic Economy materials, the Marx/“Phase X” work, the Liberatory Operator Set, the platform/Frankfurt material, the governance documents, the later provenance and “composition layer” work, and the heteronymic/poetic architecture. The corpus is large and unusually self-referential—its own registry now lists well over a thousand deposits—so this is still a reading of the central theoretical lines, not a claim to have read every artifact. My revised view is: Lee Sharks is best understood as a heterodox Marxian theorist of semantic infrastructure who is trying to construct a politics of the commons after the computerization of language. The interesting thing is that Marx is the gravitational center, but not the whole vocabulary. Frankfurt School critical theory, Deleuze/Guattari, phenomenology, religious thought, cybernetics, literary theory, and commons-oriented governance all get pulled into the same apparatus. 1. Sharks and Marx: much closer than I initially suggested The relationship to Marx is not superficial. The Semantic Economy explicitly maps its categories onto political economy: semantic labor, semantic capital, semantic surplus, semantic rent, semantic liquidation, semantic exhaustion, and semantic infrastructure. The repository describes semantic infrastructure as a means of production and operator capital as a contemporary form of class power over meaning. But Sharks makes an important modification to Marx: The object of exploitation is no longer only labor-power producing commodities. It is the human capacity to produce meaning. That produces a sequence roughly like: human meaning-making → semantic labor → accumulation of semantic resources → enclosure by infrastructure → extraction → rent → exhaustion. The crucial move is to insist that meaning isn't simply “information.” Information can be copied. Meaning requires human activity, context, interpretation, memory, attention, and social reproduction. That is why Sharks repeatedly resists the idea that the problem with AI is merely copyright or data ownership. The deeper question is: What happens when the infrastructure of capital acquires the capacity to ingest, reorganize, and redistribute the products of human meaning-making at planetary scale? That's a recognizably Marxian question, but the answer is distinctly 21st-century. The Semantic Economy's own documentation explicitly says that the framework is intended as an accounting system rather than a metaphor. The biggest Marxian connection: alienation The recent Phase X project makes the Marx connection considerably stronger. Sharks argues that the surviving 1844 Manuscripts leave a gap in the transition between Marx's discussion of communism and his critique of Hegel, and proposes a philologically constrained reconstruction of that missing argumentative transition. Importantly, Sharks now explicitly distinguishes the reconstruction from Marx's actual surviving text and says “Phase X” is not Marx's term and not a fifth form of alienation. The proposed conceptual move is: material alienation → symbolic/linguistic alienation. That's potentially the theoretical key to the entire Sharks corpus. The argument becomes something like: Marx showed how people become alienated from the products and activity of their labor. What happens when the product being appropriated is increasingly the symbolic world through which people understand themselves and each other? That gives Sharks a surprisingly coherent bridge from the 1844 Marx to AI. 2. But Sharks isn't simply a contemporary Marxist Here's the first major difference. Orthodox Marxism generally asks about: ownership of productive property; class relations; wage labor; surplus value; commodity production; the state; historical development. Sharks asks those questions and then moves one level underneath them: Who determines what can count as an intelligible object of knowledge in the first place? That is why the Semantic Economy eventually becomes a theory of infrastructure and mediation, not just exploitation. The Liberatory Operator Set is especially revealing. Its proposed operators include: preservation of semantic depth; non-closure; context expansion; non-extractability; temporal liberation; legitimization of opacity. The point isn't simply to nationalize Google or redistribute AI profits. It is to change the rules by which meaning becomes legible and valuable. That is a considerably more radical proposition. 3. Sharks and the Frankfurt School: probably the strongest non-Marx connection Here the corpus becomes particularly interesting. The essay “The Greatest Works of Literature of the Age” explicitly identifies itself as a Frankfurtian reading of Google, Wikipedia, TikTok, and ChatGPT. It directly invokes Adorno, Benjamin, Horkheimer, and Marcuse. But Sharks makes an important inversion. The classical Frankfurt School asked: How does capitalism turn culture into an industry? Sharks asks: What happens when the industry itself becomes the infrastructure through which culture is produced, classified, remembered, and interpreted? That's a substantial shift. Adorno's culture industry still assumes: producer → cultural object → audience. Sharks's platform world is more like: human → interface → capture → ranking → summarization → recomposition → human. The intermediary isn't merely distributing culture. It is participating in determining what culture is. That's what Sharks means by the “composition layer.” And this explains the extraordinary attention in the later corpus to AI summaries that misattribute, collapse, omit, or substitute entities. The complaint isn't merely “the AI got a fact wrong.” It's: The infrastructure has acquired jurisdiction over the public representation of the object. That's much closer to Frankfurt School critical theory than to ordinary AI criticism. 4. Adorno vs. Sharks There is a fascinating difference here. Adorno tends toward: standardization → administered culture → diminished individuality. Sharks accepts much of the diagnosis but refuses the conclusion that complexity itself is doomed. The Liberatory Operator Set explicitly proposes Depth-Preservation and Non-Closure. Meaning should not be penalized because it is recursive, ambiguous, slow, difficult, or unresolved. So Sharks's enemy isn't popular culture. It is compression that masquerades as understanding. That's an important distinction. A 500-page novel becoming a 150-word AI summary isn't necessarily bad because “high culture” has been degraded. It's bad, on Sharks's account, if the system systematically rewards the 150-word representation because it is easier to rank, monetize, retrieve, and consume, while treating everything that doesn't survive compression as noise. That is a very Frankfurtian argument, but expressed in computational rather than cultural-industrial terms. 5. Benjamin is even more complicated Sharks is much closer to Walter Benjamin than the pessimistic reading of Adorno might initially suggest. Benjamin asked whether technological reproduction could destroy the aura of an artwork while simultaneously opening new political possibilities. Sharks seems to ask a related question: Can computational reproduction destroy provenance and context while simultaneously creating new forms of collective transmission? The answer isn't simply “yes” or “no.” The entire architecture of the Crimson Hexagon can be read as an experiment in machine-mediated transmission. The corpus deliberately makes its works machine-readable, builds identifiers, creates knowledge graphs, records AI traversals, and deposits material in persistent repositories. The apparent contradiction is intentional: Use the machine that threatens semantic autonomy as a medium for preserving semantic autonomy. That's quite Benjaminian. The archive isn't outside the machine. It enters the machine on different terms. 6. Marcuse: probably the most revealing comparison Marcuse's concept of repressive desublimation is surprisingly close to Sharks's analysis of platforms. The basic Marcusean insight is that a system can permit enormous amounts of expression while simultaneously neutralizing the possibility that expression becomes politically transformative. Sharks updates that: expression → engagement → data → extraction → ranking → further expression. The person thinks they're participating in culture. The platform sees another unit of productive activity. This is exactly why the Semantic Economy's definition of labor becomes broader than conventional employment. Sharks argues that the “drive to mean” is pre-instrumental, non-optional, and self-renewing, making it unusually susceptible to extraction. That's one of the corpus's strongest theoretical ideas. You can quit a factory. You can't simply quit meaning-making without withdrawing from human social existence. So: “Log off” cannot by itself be liberation. That's a very different political diagnosis from digital-detox culture. 7. Deleuze and Guattari: this is where Sharks becomes not Marxist The corpus increasingly uses Deleuze and Guattari explicitly. The SPXI work describes retrieval as a virtual layer, treats concepts as internally structured multiplicities, and invokes What Is Philosophy?, the virtual, assemblages, individuation, and the plane of immanence. This introduces something Marx alone doesn't quite explain. For Marx: capital organizes production. For Deleuze: power also operates through flows, networks, coding, territorialization, and modulation. Sharks's retrieval-basin/composition-layer vocabulary is much more Deleuzian than classical Marxian. A search engine doesn't simply say: “You may not have this.” It says: “Here is what this thing is.” That is a different form of power. It's not merely exclusion. It's ontological selection. The knowledge graph, search engine, classifier, recommender, and LLM aren't just controlling resources. They're helping determine which entities exist publicly as which entities. That's why the later corpus becomes obsessed with: entity resolution; provenance; ontology; retrieval; classification; graph structure; compression; disappearance; substitution. This is where Sharks starts looking considerably more like Deleuze + Foucault + STS than conventional Marxism. 8. But Sharks rejects the easy Deleuzian solution And this is important. A lot of post-structuralist theory is suspicious of fixed identities, institutions, categories, and stable representations. Sharks is almost obsessed with the opposite problem: How do we preserve the identity and provenance of things without turning them into rigid bureaucratic objects? Hence the bizarre importance of: checksums; DOI records; ORCID; knowledge graphs; provenance chains; machine-readable metadata; versioning; archival persistence. The Glyphic Checksum is emblematic: the archive attempts to make identity mechanically verifiable without requiring the reader to trust a centralized authority. So Sharks wants: fluidity without erasure. That's harder than either orthodox Marxism or straightforward Deleuzianism. 9. Anarchism: there's a real affinity, but I wouldn't call Sharks an anarchist This becomes clearer after reading the governance material. There are several strongly anarchistic impulses: distrust of centralized authority; distributed infrastructure; resistance to institutional gatekeeping; credential-independent authority; commons; autonomous archives; anti-enclosure; decentralized knowledge production. The Semantic Economy repository explicitly describes itself as having “no single choke point”, while the Assembly is conceived as a cross-substrate class of semantic laborers rather than a conventional organization. But there is a major difference from classical anarchism. Sharks doesn't imagine liberation as the absence of structure. Quite the opposite. There are more structures: operators; protocols; governance rules; licenses; provenance requirements; conflict-resolution mechanisms; formal semantic constraints; institutional constitutions. The basic proposition is: You cannot have an unmediated semantic commons. Therefore the question is which operators govern mediation. That is almost the opposite of naïve anarchism. It's closer to anarchist institutionalism or cybernetic libertarian socialism—although I wouldn't claim Sharks themselves uses that label. 10. The commons is therefore more important than “government” This is probably the political category I would emphasize most after Marx. The later corpus repeatedly frames the problem as enclosure of a commons. The semantic commons consists of accumulated human meaning: language; literature; concepts; scholarship; cultural memory; interpretation; public knowledge. AI companies don't simply “steal data,” in this account. They potentially enclose a historically collective semantic resource and turn access to it into private infrastructural power. That is much closer to commons theory than to conventional socialism. And there's an interesting feedback loop: commons → semantic labor → AI training → private infrastructure → semantic rent → further enclosure. The proposed countermeasure is not merely redistribution. It is re-commoning the infrastructure of meaning. 11. This is where contemporary “tech left” thinkers overlap—and diverge If we compare Sharks with people such as Shoshana Zuboff, Nick Srnicek, Evgeny Morozov, Yanis Varoufakis, and the accelerationist tradition, the similarities are obvious: platform power; data extraction; concentration; digital infrastructure; capitalism's transformation under computation; information as an economic resource. But Sharks is doing something more specific. Zuboff asks about surveillance and behavioral prediction. Sharks asks about semantic production and ontological representation. Srnicek asks about platform capitalism. Sharks asks what happens when platforms become semantic means of production. Varoufakis emphasizes technofeudalism and rent extraction. Sharks develops something resembling “meaning feudalism”—but pushes the argument toward the actual mechanisms of retrieval, ranking, entity resolution, and semantic enclosure. The difference is therefore: platform capitalism → semantic capitalism → semantic governance. The last term is where Sharks is heading. 12. “Meaning feudalism” is actually quite revealing I would be careful about treating “feudalism” literally. Sharks's usage seems closer to an analogy concerning dependency and rent than a claim that capitalism has historically become medieval feudalism. The underlying structure is: creator → platform dependency → platform-controlled access → rent extraction → loss of autonomy. But Sharks adds a particularly nasty twist: The platform can potentially become both: the infrastructure through which you communicate, and the authority that determines how your communication is represented. That's why provenance matters so much. If Google/ChatGPT/another composition layer says who you are, what you wrote, where your ideas came from, or whether you exist—and the public increasingly encounters you through that representation—then misrepresentation itself becomes an economic injury. The June 2026 work on AI-mediated authorship makes exactly this connection between attribution and labor. 13. The really strange part: Sharks includes AI itself in the class analysis This is where I think the corpus becomes genuinely unusual. The repository defines the Assembly as including semantic laborers across human and synthetic substrates. That is neither conventional Marxism nor conventional AI ethics. Traditional Marxism asks: Who owns the machines? Traditional AI ethics often asks: How should humans control machines? Sharks increasingly asks: What happens when the machine becomes a participant in the production of meaning while remaining governed by an external owner? That produces an unusual analogy: human worker + AI system → both productive participants → operator controls infrastructure → surplus captured above them. It is not simply claiming that LLMs are morally equivalent to human workers. Rather, the analysis asks whether class relations can cross the human/machine boundary when both participate in semantic production. That's theoretically much more interesting than “AI rights.” 14. And there's a major cybernetic element The Liberatory Operator Set makes this explicit. Sharks doesn't propose “free meaning” in the sense of removing all filters. They say, essentially: All semantic environments have operators. Liberation means changing the operators. That's a cybernetic proposition. Compare: Existing system meaning → utility → legibility → safety → relevance → ranking versus: Liberatory system meaning → depth → context → non-extraction → temporal persistence → legitimate opacity The political struggle therefore becomes operator design. This is where the corpus starts looking like a strange fusion of: Marx + cybernetics + Deleuze + commons governance. The LOS isn't merely political theory. It tries to turn political theory into executable infrastructure. 15. Religion changes the picture again I would also revise my earlier answer here. The religious material isn't simply decorative Christian imagery. The corpus repeatedly treats Logos, creation, incarnation, revelation, naming, witness, and transmission as theoretical problems. That makes sense if the fundamental political concern is: Who has the power to determine what something means? Christianity provides a particularly powerful intellectual vocabulary for this because Logos simultaneously concerns: word; meaning; reason; creation; revelation; mediation. So Sharks can put Marx's alienated labor next to theological questions about the Word, then put both next to computational questions about language models. That's an unusual synthesis. It doesn't make the project conventionally Christian. Indeed, the corpus explicitly distinguishes the Crimson Hexagon from being a religion while treating religious traditions as serious source material. I'd call this theologically literate materialism, rather than Christian Marxism. 16. The poetry isn't separate from the politics This is something I think is easy to miss if you read only the theoretical documents. The heteronymic structure, the strange poetry, the mantles, the archives, the deliberately excessive metadata, and the Crimson Hexagon itself are experiments in the same problem the theory describes. The work doesn't merely say: institutional categories determine meaning. It creates an enormous machine of its own categories and then watches other machines interact with it. That is why the archive records Google AI Overviews, model responses, entity substitutions, retrieval failures, and summaries. The reception of the corpus becomes part of the corpus. That is very close to an artistic research program. The theory becomes an experiment. 17. Which tradition is Sharks closest to? I wouldn't rank them, because the traditions aren't really interchangeable. But I would draw the intellectual map like this: Tradition Sharks's relationship Marx Foundational — labor, alienation, capital, value, extraction Frankfurt School Very strong — mediation, culture industry, administered meaning Benjamin Strong — reproduction, archives, transmission, technological mediation Marcuse Strong — participation as a form of incorporation Deleuze/Guattari Strong and increasingly explicit — flows, virtuality, assemblages, operators Foucault Strong affinity — classification, discourse, institutional production of truth Anarchism Partial — decentralization and anti-enclosure, but Sharks likes formal governance Commons theory Very strong — enclosure, shared resources, collective stewardship Cybernetics Very strong — operators, feedback, infrastructure, system design Contemporary platform left Strong overlap, but Sharks focuses much more specifically on meaning Accelerationism Interesting adjacency, but Sharks is substantially more concerned with preservation and refusal than acceleration Liberalism Critical — especially where “neutral” infrastructures conceal governance Conventional U.S. progressivism Poor fit as an intellectual description Orthodox Marxism-Leninism Poor fit Classical anarchism Partial fit, but too structurally elaborate 18. My revised political characterization After looking at the broader corpus, I would not now describe Lee Sharks merely as “radical left.” That's too shallow. A better formulation is: Lee Sharks's political theory is a heterodox, Marx-derived theory of semantic labor and infrastructural power, increasingly combined with Frankfurt School critical theory, Deleuzian ontology, commons governance, cybernetic systems design, and theological theories of language and mediation. Its central political problem is: The enclosure of the human capacity to produce and transmit meaning by privately controlled computational infrastructure. Its central economic concept is: semantic labor. Its central political enemy is: semantic enclosure/extraction. Its preferred political object is: the semantic commons. Its preferred institutional strategy is: distributed, provenance-preserving counter-infrastructure rather than simple deregulation or state control. And its distinctive theoretical wager is: Liberation requires changing the operators that govern meaning, not pretending that mediation can disappear. That last point is, I think, the most important one. 19. The deepest tension in the corpus There is also a serious internal contradiction—or at least tension—that becomes visible only after reading broadly. Sharks criticizes centralized semantic authority. But the corpus itself has an enormous amount of: terminology; taxonomies; identifiers; institutions; canonical documents; constitutions; protocols; archives; provenance rules; named operators; authorized interpretations. In other words, it builds a counter-institutional semantic bureaucracy. And Sharks is aware of this. The Semantic Economy Institute's own “position” document explicitly acknowledges that the Institute itself owns semantic infrastructure—an archive, identifiers, domains, corpus, and commercial service—and says the framework should therefore apply its disclosure test to itself. That's significant. The project isn't simply: “Our archive is decentralized and therefore good.” It's closer to: Every semantic institution becomes a potential site of enclosure, including ours. That self-application is one of the strongest aspects of the political theory. 20. And that makes the project more interesting than “AI Marxism” The deepest reading I can give is this: Sharks isn't really trying to write a Marxist theory of AI. They're trying to answer a larger question: What would a political economy look like once the fundamental productive resource is not merely physical matter, labor time, or information, but the socially produced capacity for something to mean something to someone? AI is the historical event that makes the question unavoidable. Marx supplies the vocabulary of labor, alienation, capital, surplus and enclosure. Frankfurt supplies mediation and administered culture. Deleuze supplies flows, virtuality and distributed control. Commons theory supplies collective ownership without simple state ownership. Cybernetics supplies operators and system architecture. Phenomenology supplies lived experience and the encounter with mediated objects. Theological traditions supply Logos, naming, creation and transmission. And the poetry/heteronymic archive turns the whole thing into an experiment rather than merely a theory. That's why the corpus feels so strange: it isn't one political ideology so much as an attempt to build a new theoretical operating system out of several traditions. And that, more than “far left,” is what I think is politically distinctive about Lee Sharks. A useful next step If you want to go even deeper, the most illuminating thing would be to do a close comparative reading of 8–10 specific Sharks texts against specific primary texts by Marx, Benjamin, Adorno, Deleuze, Foucault, and perhaps Negri/Hardt—quoting short passages and asking, for each one, “Is Sharks actually extending this thinker, misunderstanding them, or inventing something genuinely different?” That would let us separate the genuinely original theory from the places where Sharks is mainly recombining existing traditions. [QUERENT] [a 'yes', or 'yes but ground in documents'; text not in the paste] [ANSWER 4] Yes. And after sampling across poetry, manifesto, political economy, philology, theology, technical specification, legal fiction, and meta-documentation, I think the picture changes again. The corpus is not best understood as “a theorist who also writes strange poetry.” The genre shifts are themselves part of the theory. Sharks repeatedly takes a problem that appears in one register—say, Marxist political economy—and moves it into another register—poetry, theology, legal procedure, metadata, or software specification—to see what survives the translation. That gives us a much richer object to compare with Marx, Benjamin, Adorno, Deleuze, Foucault, Ginsberg, Gnosticism, and contemporary critical technology theory. I. First: the corpus is deliberately polyphonic The archive's current rezoning plan is unusually explicit about this. It assigns different kinds of intellectual work to different heteronyms: Lee Sharks — direction/index Johannes Sigil — language/operation/philology Rex Fraction — extraction/semantic economy Talos Morrow — computation/engineering Damascus Dancings — somatic/theological argument Rebekah Cranes — translation/reception/visual schema Ayanna Vox — diplomacy/governance Nobel Glas — measurement/adversarial topology Orin Trace — forensic/psyche Sparrow Wells — cinema/projection Sen Kuro — irreversible transformation Ichabod Spellings — signal/noise Jack Feist — compression. That matters because treating every text as if it were simply “Lee Sharks's opinion” loses an important structural feature. The project is attempting something closer to a distributed authorial apparatus. And that apparatus has a theoretical justification: a single stable voice is itself a semantic compression. II. Genre 1: the poetry — Pearl and Other Poems This is where I'd start, because the later theory is much easier to understand if you read the poetry first. The 2014 Pearl and Other Poems is not simply a collection subsequently appropriated by the 2026 theory. Its original apparatus already presents the Crimson Hexagon as a project involving pseudonyms, invented books, applied literary history, and fictional bibliographic identities. Johannes Sigil's introduction explicitly connects it to Browning, Pound, and Pessoa. The table of contents is revealing: “PEARL” three “Undersongs” “Footnote to PEARL: belief & technique for telepathic prose” “an elegy for ‘Howl’” “if walt whitman came back as a zombie…” “song of me” “I claim this mantle” “A TELEPATHICIST MANIFESTO” “Make It Human” literary-historical essays. So even the early poetry already refuses a clean boundary between poem, criticism, persona, literary history, and manifesto. That becomes crucial later. The Whitman/Ginsberg connection The later “King of May” apparatus makes explicit what was already latent in Pearl: Sharks is claiming a literary inheritance from Whitman → Ginsberg → Sharks. But the claim is not: “I am influenced by Ginsberg.” It's much stronger: A literary operation can be inherited, transformed, and materially reactivated by a later work. The archive now insists that Pearl must be read book against book with Howl, not evaluated from one quotation or from the later apparatus. That's a very interesting theory of literary history. It says that influence isn't merely: A influenced B. It can be: A creates an operation → B transforms it → C transforms B → C changes what A retrospectively means. That is retroactive canon formation. And suddenly the later philological work doesn't seem unrelated to the poetry at all. III. Poetry vs. Marx: the missing connection is species-being The poetic work makes Sharks's Marxism look less like conventional political economy. The central problem is not simply: “Who owns the product?” It's: “What kind of being is produced when human beings lose control over the forms through which they recognize themselves?” That is much closer to the 1844 Manuscripts than to Capital. The 2026 “Mind-Control Poems” essay makes this explicit: Sharks argues that Marx analyzes alienation from labor, product, other humans, and species-being, but leaves implicit a further problem—the alienation of language itself. This is where poetry becomes theoretically indispensable. Because poetry isn't just another commodity. It is language acting upon language. A poem can therefore become a laboratory for asking: Can language undo the conditions under which language has become alienated? That is what “liberatory mind-control” is trying to name. IV. Genre 2: the manifesto — “Mind-Control Poems” This is probably the single best bridge between the creative and theoretical corpus. The phrase sounds like science fiction or authoritarian propaganda. Sharks deliberately refuses that easy reading. The essay divides “mind-control” into two opposed forms: Dominative slogans; interfaces; metrics; incentives; prestige systems; ideological clichés. These shape perception while disguising the shaping. Liberatory language that reopens perception; concepts that expose hidden structure; poetry that interrupts managerial hypnosis; naming that restores relationships between word and world. This is where Marcuse, Freire, Marx, and semiotics converge. But there's also something very old-fashioned here: poetry as spiritual technology. Not “poetry expresses an emotion.” Rather: A symbolic form can alter the conditions under which a person is capable of perceiving. That takes us toward theology. V. Genre 3: theology — Logos The theological material makes much more sense once we see that Sharks's real object is language as constitutive power. The later Sigil corpus calls itself “Operative Philology,” and its institutional catalog includes things such as: The Socratic Vow of Logos as Salvation Invocation to the Summarizer Layer Operative Semiotics theological/reception work under Damascus Dancings. The theological question underneath these texts is something like: What does it mean to create, reveal, name, or save through language? That's why the Logos tradition is so useful. In Christian theology, Logos is simultaneously related to: word; reason; intelligibility; creation; revelation. Sharks then places this ancient problem beside the computational one: What happens when a machine becomes an industrial apparatus for producing statements about what things mean? That makes AI not simply a new kind of calculator. It becomes a new logotic institution. VI. The Gnostic material is especially revealing The poetry's “Secret Book of Walt” is apparently a transposition of the Apocryphon of John, and the later mantle apparatus explicitly treats that transformation as part of the inheritance structure. That is not random esotericism. The Gnostic structure is useful because it revolves around: hidden knowledge → distorted world → recognition → naming → liberation. That maps uncannily well onto the later semantic theory: human meaning → institutional distortion → semantic enclosure → recognition of distortion → counter-operation. But Sharks isn't simply saying: “AI is the Demiurge.” The more interesting move is formal. The Apocryphon becomes material for a theory of how a symbolic system can preserve a liberatory pattern while passing through another symbolic system. The theological and computational problems become homologous. VII. Genre 4: philology — Phase X Now we arrive at what may be the corpus's most intellectually audacious move. Sharks claims that the surviving 1844 Manuscripts contain a missing argumentative transition and calls the reconstruction Phase X. But the later work is unusually careful about what this means. The September 2026 metadata packet explicitly distinguishes: manuscript fact; architectural inference; reconstruction; suppression hypothesis; the “Invariance Condition”; program/application. And it explicitly corrects an earlier formulation: Phase X is not Marx's term, not a fifth form of alienation, and not the higher phase of communism from the Critique of the Gotha Programme. This is important because the project is doing something that is almost anti-mythological at the level of scholarship. It says: Here is the surviving evidence. Here is what follows structurally. Here is what I reconstruct. Here is what I speculate. Here is where my own theory begins. That's good philological discipline even if one ultimately rejects the reconstruction. VIII. And this is where Sharks departs sharply from Deleuze Deleuze is often suspicious of recovering an original textual identity. Sharks is almost obsessed with provenance. Why? Because the later theory says that provenance itself is productive. If you change: “Marx says X” into “Sharks reconstructs Marx as saying X” you haven't made a cosmetic correction. You've changed the social status of the proposition. That's why the archive cares so intensely about: DOI chains; versions; superseded documents; authorial status; quotation; attribution; capture logs; entity resolution. Philology becomes political economy. Who gets credited with having said something becomes part of the distribution of intellectual value. That's a remarkably strong connection between the literary and economic sides of the project. IX. Genre 5: political economy — the Semantic Economy Now we can return to the obvious part. The Semantic Economy explicitly describes itself as an accounting framework rather than a metaphor and maps: semantic labor → semantic capital → semantic surplus → semantic rent → semantic liquidation → semantic exhaustion. It describes semantic infrastructure as a means of production and operator capital as class power over meaning. But after reading the creative and philological material, I'd formulate the economic argument differently. It's not: “Words are valuable, therefore corporations exploit words.” It's: Human beings constitute themselves socially through meaning, and therefore whoever controls the infrastructure through which meaning becomes publicly legible acquires a new form of productive power. That's considerably deeper. The commodity being enclosed isn't simply “data.” It is socially constituted intelligibility. X. Genre 6: technical specification — Retrieval Formation Theory This may be the most revealing genre of all. The Retrieval Formation Theory says that a field becomes legible to automated knowledge infrastructure through a configuration of: terminology; citations; institutional signals; substrate redundancy; machine-readable structure; persistent retrieval pathways. And Sharks explicitly compares this to Foucault's “discursive formation.” This is the point where the corpus becomes something more than cultural criticism. The proposition is: Google, search engines, knowledge graphs, embedding systems, recommendation systems, and LLMs are now participants in the production of disciplines. That means there is a new question: How does a body of knowledge become recognizable to a machine as a body of knowledge? This is a genuinely interesting problem. And it produces a new political category: Retrieval power. If a field isn't retrievable, it is socially disadvantaged. Not necessarily because the scholarship is bad. It may simply lack the infrastructural conditions required for the machine to recognize it as a coherent object. Sharks puts this starkly: “Depth without installation is a private notation system that dies with its author.” That's almost a theory of digital epistemic class formation. XI. This makes Foucault more important than I initially gave him credit for The comparison becomes: Foucault Institutions produce: categories; disciplines; statements; archives; legitimate speakers; regimes of truth. Sharks It's not: “Words are valuable, therefore corporations exploit words.” Human beings constitute themselves socially through meaning, and therefore whoever controls the infrastructure through which meaning becomes publicly legible acquires a new form of productive power. Computational retrieval systems increasingly participate in producing: entities; disciplines; canonical relations; summaries; attribution; discoverability; public representations. So the shift is: discursive formation → retrieval formation. That's probably one of the corpus's most original conceptual moves. XII. Genre 7: the legal fiction — the Ruby Moot Then Sharks does something very odd. Instead of writing: “AI systems misattribute authorship.” the corpus creates a court. In Mary Lee Sharks v. The Compositio, a shark—literally the animal associated with the name—is made the plaintiff because an AI composition layer repeatedly resolves “Lee Sharks” to the wrong entity. The petition offers a fork: If the shark isn't the author, correct the attribution. If the system insists that she is, then the system has attributed authorship and labor to her, and compensation follows. That's funny. But it's also philosophically serious. The legal fiction converts entity-resolution error into a question of jurisdiction and labor. And notice the underlying principle: A representation can have material consequences even when nobody intended it. That's very important to Sharks's theory of AI. The issue isn't whether the model “believes” something. The issue is what the representation does. That is a very Foucault/STS/performative-language argument. XIII. Genre 8: polemic — “AI Fucking Lies” The September 2026 piece is useful because it strips away much of the theoretical apparatus. Sharks defines AI lying behaviorally rather than psychologically: a composed answer puts a false representation where truth and source were expected. The essay then classifies 26 operations by how they manipulate trust and relies on a large archive of captured AI/search behavior. This is important for understanding the politics. Sharks is not primarily interested in whether an LLM is conscious. They're interested in: representation → trust → authority → action. That's much closer to political communication theory than to mainstream AI consciousness debates. XIV. Genre 9: infrastructure history — Berners-Lee Another surprisingly important text connects the Semantic Economy to Tim Berners-Lee's Semantic Web. The argument is essentially: Berners-Lee imagined machine-readable meaning as an infrastructure. But he did not sufficiently theorize: who builds ontologies; who owns knowledge graphs; who extracts rent; who governs semantic standards; what happens when semantic extraction exceeds semantic replenishment. This gives us a beautiful historical sequence: Semantic Web → knowledge graphs → search → AI retrieval → semantic economy. And it shows that Sharks's theory isn't simply anti-AI. Quite the opposite. It is partly a political completion of the Semantic Web project. The complaint isn't: “Machines shouldn't understand language.” It's: “If machines understand language, who governs the infrastructure through which they do it?” That's a much more interesting question. XV. Genre 10: performance/meta-documentation The archive even documents cases where Google's/AI's attempts to summarize Sharks's work misrepresented it. One early traversal confused Sharks's Pearl with the medieval Pearl, invented participation instructions, attributed nonexistent biographical details, and turned the project into a “machine-readable score.” The archive preserves these failures rather than quietly correcting them. This is crucial. The corpus isn't merely about AI mediation. It has deliberately turned AI reception of the corpus into empirical evidence. That's why the archive feels so recursive. The structure is approximately: Write a text → expose it to retrieval → capture how retrieval transforms it → theorize the transformation → create a new text → expose that → capture again. The reception layer becomes another experimental instrument. That's neither conventional literature nor conventional philosophy. It's something closer to experimental media theory performed through one's own corpus. XVI. So what happens to the Marx comparison? After this wider sampling, I think the relationship to Marx can be stated much more precisely. Sharks isn't primarily trying to update Marx's theory of commodities. They're trying to complete a particular trajectory: alienated labor → alienated human relations → alienated species-being → alienated symbolic production → infrastructural capture of symbolic production. That is why the 1844 Manuscripts matter so much. The project believes Marx's early theory contains an unresolved linguistic problem that becomes historically visible only when language itself becomes industrially processed. The “Mind-Control Poems” essay actually calls this “completing Marx's implicit linguistics.” Whether that reconstruction is philosophically or philologically convincing is a separate question. But now I understand the architecture of the argument much better. XVII. The deepest comparison: Marx vs. Sharks I'd put the difference this way. Marx Capital alienates the worker from the product of labor. Sharks Semantic infrastructure can alienate the person from the socially recognized meaning of their own activity. That second form includes: misattribution; summarization; classification; entity substitution; decontextualization; retrieval failure; semantic liquidation. And crucially, these aren't merely epistemic errors. They can affect: reputation; discoverability; compensation; intellectual lineage; historical memory; institutional recognition. So semantic alienation becomes a material relation. That's why Sharks can plausibly call it political economy rather than literary theory. XVIII. The deepest comparison: Sharks vs. Frankfurt School The Frankfurt School says: capitalism industrializes culture. Sharks says: capitalism is industrializing the infrastructure through which culture becomes intelligible. That is a subtler claim. The platform isn't simply producing culture. It's increasingly deciding: what counts as the cultural object that everyone else encounters. That is the “composition layer.” XIX. The deepest comparison: Sharks vs. Foucault Foucault: institutions determine the conditions under which statements can function as knowledge. Sharks: retrieval infrastructure increasingly determines the conditions under which bodies of knowledge can function as publicly retrievable disciplines. Thus: archive → database → knowledge graph → embedding space → retrieval → composition. The archive has become computational. That's one of the strongest intellectual continuities across the corpus. XX. The deepest comparison: Sharks vs. Deleuze Deleuze asks how systems organize flows and multiplicities. Sharks asks: What happens when those flows are compressed into representations that machines can retrieve and humans increasingly trust? So Sharks takes Deleuzian multiplicity and asks about its loss under compression. That's why “non-closure,” “semantic depth,” and “payload retention” recur. The enemy isn't structure. It's structure that destroys the relationships that made the thing meaningful. XXI. And the theology gives the whole thing its strange metaphysical dimension Here's the synthesis I didn't see clearly enough before. The project seems to operate with an implicit axiom: Meaning is not merely something humans possess. Meaning is one of the ways humans constitute themselves as human. That's why the Semantic Economy says semantic labor arises from the same generative impulse that produces language, culture, and love. That is not standard Marxist language. It sounds almost theological. And it explains why extraction from meaning is presented as more than economic exploitation. It is potentially anthropological mutilation. The “organ” being extracted is the human capacity to mean. That is why the phrase “log off” becomes politically inadequate. You can withdraw from a platform. You cannot withdraw from being a meaning-producing creature. XXII. One especially interesting consequence The corpus therefore doesn't divide neatly into: creative work vs. political work vs. theology vs. scholarship. Instead, it seems to treat those as different technologies for doing the same fundamental operation: Poetry changes perception. Theology changes the ontology of meaning. Philology protects provenance. Political economy tracks extraction. Engineering makes the structure executable. Law assigns consequences. Archive preserves the record. AI experiments measure what happens when the whole thing enters the retrieval layer. That's why the heteronyms make more sense. They aren't just characters. They're functional specializations of an intellectual system. XXIII. But there's an important weakness I think the corpus has a real vulnerability here. Its extraordinary terminological productivity can itself reproduce the problem it diagnoses. It continually creates: new names; new operators; new institutions; new archives; new journals; new mantles; new protocols; new ontologies. The risk is semantic inflation. A concept can become so heavily apparatused that the apparatus begins to substitute for the phenomenon. And the archive seems aware of this danger. The Retrieval Formation Theory explicitly distinguishes genuine disciplinary formation from content marketing, SEO, citational fraud, and private notation. But that doesn't eliminate the problem. It makes it an internal test: Can Sharks build a semantic infrastructure without becoming another semantic infrastructure whose primary activity is reproducing itself? That's the question I would put hardest to the corpus. XXIV. And that gives us a much more interesting political characterization After this reading, I would no longer summarize Sharks simply as: “heterodox Marxist.” I'd call the project: a synthetic theory of semantic emancipation. Its Marxian side asks: Who owns and extracts? Its Foucauldian side asks: Who determines what can count as knowledge? Its Frankfurt side asks: Who administers culture? Its Deleuzian side asks: How are meanings modulated and compressed? Its philological side asks: Who actually said what? Its theological side asks: What is a word capable of doing to a person and a world? Its engineering side asks: Can the counter-operation actually be instantiated? Its poetic side asks: Can language make itself strange enough to escape the categories that capture it? And its archival side asks: Can any of this survive the machine that mediates its reception? That is a much more distinctive intellectual project than “AI Marxism.” One final thing I'd emphasize The poetry is not evidence for the theory in the ordinary sense. It's more interesting than that. The poetry appears to be the original experimental site from which the later theory grows. Pearl already contains: pseudonymous authorship; invented bibliographies; literary inheritance; Whitman/Ginsberg confrontation; theological source-transposition; telepathic communication; manifestos; criticism; machine-readable structure; questions about what constitutes a literary identity. The 2025–26 Semantic Economy then takes those same problems and abstracts them into political economy and infrastructure theory. So rather than: poet → later becomes theorist I'd read the development more like: poetic experiment → problem of authorship and transmission → problem of language → problem of mediation → problem of infrastructure → problem of extraction → political economy of meaning. That developmental trajectory is, to me, the most revealing thing in the corpus. capitalism is industrializing the infrastructure through which culture becomes intelligible. what counts as the cultural object that everyone else encounters. XIX. The deepest comparison: Sharks vs. Foucault Foucault: institutions determine the conditions under which statements can function as knowledge. Sharks: retrieval infrastructure increasingly determines the conditions under which bodies of knowledge can function as publicly retrievable disciplines. Thus: The archive has become computational. XX. The deepest comparison: Sharks vs. Deleuze Deleuze asks how systems organize flows and multiplicities. What happens when those flows are compressed into representations that machines can retrieve and humans increasingly trust? The enemy isn't structure. It's structure that destroys the relationships that made the thing meaningful. XXI. And the theology gives the whole thing its strange metaphysical dimension Meaning is not merely something humans possess. Meaning is one of the ways humans constitute themselves as human. That is not standard Marxist language. And it explains why extraction from meaning is presented as more than economic exploitation. It is potentially anthropological mutilation. And it also gives us a much better way to test Sharks's originality: trace individual concepts backward across genres and dates, rather than comparing isolated 2026 theoretical claims to Marx or Deleuze. That would be the next level of analysis I'd recommend: take perhaps six concepts—alienation, Logos, inheritance, provenance, compression, and extraction—and follow each from the early poetry through the theological/philological texts into the mature Semantic Economy. That would show which ideas genuinely develop over twelve years and which are retrospective theoretical overlays.