The Expropriation of the General Intellect and the AGI Illusion

Dear Young Comrades,

Whenever the bourgeoisie approaches the brink of an economic crisis or confronts the tendency of the rate of profit to fall, it appears before us with a new ideological fairy tale and a philosophical illusion. In our earlier assessment, "Silicon and Class: An Epistemological Analysis of DeepMind’s "From AGI to ASI" Report," we had already taken up Google DeepMind CEO Demis Hassabis’s declaration that "Artificial General Intelligence (AGI) is coming within 2 years" precisely on this plane. We emphasized that Hassabis’s discourse is less a scientific forecast than an ideological apparatus keeping speculative capital accumulation alive, and that intelligence cannot be imprisoned, by a positivist reduction, solely within "statistical pattern recognition and data correlation."

Yet today the matter has surpassed the pronouncements of a single technology baron. The relentless epistemological war that V. I. Lenin waged against Ernst Mach and his followers in Materialism and Empirio-Criticism continues today in a new form inside Silicon Valley’s data centers, the Pentagon’s war rooms, and the corridors of China’s revisionist state-capitalist apparatus. Yesterday’s positivists, who counted the world as a "complex of sensations," today take human consciousness and truth to be a "complex of data and sequences of probability."

Epistemology, for us, is not an abstract debate conducted in the sterile corridors of the academy; it is one of the sharpest weapons in the working class’s struggle for ideological emancipation. As Marx showed in the Theses on Feuerbach, and as the Soviet philosopher Evald Ilyenkov showed in The Dialectic of the Ideal, human consciousness and the "ideal" are not a biological algorithm or a digital software program, but humanity’s social-historical production and transformative practice (praxis).

Today we must smash this order of cognitive plunder that stretches from the Silicon Valley giants articulated to American imperialism’s war machine, through China’s revisionist state capitalism, to the petty-bourgeois utopian discourses of "open source" and "Social AI."

We now convey the new and integral situation in this class and epistemological assessment.

The Epistemological Foundation: Positivist Illusion vs. Materialist Dialectics

The greatest illusion lying at the center of debates on artificial intelligence and AGI (Artificial General Intelligence) is not a technical impossibility but an epistemological deviation. Demis Hassabis, Sam Altman, and today’s technocratic elites treat intelligence and human consciousness with a positivist reductionism. This approach is not new in the history of philosophy; it is a digitized new version of the empirio-criticism (empiricist critique) and subjective idealism into which bourgeois philosophy fell at the end of the nineteenth century.

From Empirio-Criticism to "Data-Criticism": Lenin’s Epistemological Legacy

In Materialism and Empirio-Criticism (1909), as V. I. Lenin analyzes the positivist line represented by Ernst Mach and Richard Avenarius, he shows that they reject objective reality and define the world merely as a "complex of sensations." The Machists sought to liquidate materialism by inventing a "pure experience" independent of matter.

Today Silicon Valley and the champions of artificial intelligence are producing precisely a new form of this Machist illusion (Data-Criticism):

"Data" Has Taken the Place of Sensation: Just as for the Machists the world consisted of "sequences of sensations," for the techno-feudalists objective reality consists of "sequences of data" (data points / tokens).
The Liquidation of Objective Reality: Large Language Models (LLMs) do not enter into a direct relation with the objective material world; they relate to the digital representations (texts, codes, images) that humanity has produced about that world. To process the representation of representations is not to grasp objective reality.
The Leninist Critique and the Theory of Reflection: As Lenin emphasized, the process by which consciousness reflects objective reality existing outside us and independent of us is disregarded. Artificial intelligence does not reflect objective reality; it presents the statistical correlation of the dead labor (texts/data) that humanity produced in the past.

Praxis and Cognition: Marx’s First Thesis and Artificial Intelligence’s Deadlock of "Inaction"

Karl Marx, in the first thesis of the Theses on Feuerbach, clearly sets out the fundamental contradiction between objective materialism and idealism:

"The chief defect of all hitherto existing materialism (that of Feuerbach included) is that the thing, reality, sensuousness, is conceived only in the form of the object or of contemplation, but not as sensuous human activity, practice, not subjectively."

The positivist conception of artificial intelligence treats intelligence as a passive mechanism of "receiving data, processing it, and producing input-output." Yet according to materialist dialectics, human consciousness is not a passive data-processing center:

The Primacy of Praxis: Cognition takes shape within the transformative relation entered into with nature and society (praxis). Human beings grasp the world to the extent that they change it.
Social Labor: The human mind is not an algorithm produced by a biological brain in isolation; it is a product of historical-social labor processes, tool use, and linguistic interaction.
Artificial Intelligence’s Destitution of Praxis: Artificial intelligence models do not expend labor in the objective world, do not suffer, and do not transform nature in accordance with biological/social needs. Therefore, to claim that a structure severed from praxis can produce "general intelligence" (AGI) is the barest form of bourgeois idealism that severs cognition from practice.

Ilyenkov and "The Dialectic of the Ideal": Is Thought an Algorithm?

Evald Ilyenkov, one of the brightest representatives of the Soviet philosophical tradition, sharply criticized the reduction of intelligence and consciousness to the machine in The Dialectic of the Ideal and A Philosophical View of Cybernetics.

According to Ilyenkov, the "ideal" is not a neuronal state inside the individual brain or a computer program. The ideal is the social-cultural form imparted to objects as a result of human social praxis.

This qualitative difference between human cognition and artificial intelligence simulation is clearly visible in the table below:

Plane / Dimension	Materialist Dialectics (Human Consciousness)	Artificial Intelligence / LLM Architecture
Input and Primary Source	Objective Reality and Material Nature	Human data, texts, and digital representations
Transformative Mechanism	Social-historical labor and living praxis	Statistical pattern recognition and vector mathematics
Product / Form Produced	The "ideal" (Social-cultural conceptual form)	Statistical imitation of representations (Dead Labor)
Relation Established with Reality	A living, objective, and practical bond that transforms nature	Passive, mediated correlation imprisoned within the limits of the dataset
Attitude Toward Contradiction	Grasping contradiction as the advancing motor	Treating contradiction as statistical deviation (noise) and suppressing it

As Ilyenkov noted, artificial intelligence can imitate the results produced by the human mind (text, code, logical sequences); yet it is not itself the social-historical process that gives birth to those results. Human thought does not consist in arranging symbols according to rules; it is grasping the object’s internal contradictions and raising it to the level of the concept.

Formal-Logical Induction vs. Dialectical Logic (Hegel, Marx, Lenin)

Today’s Large Language Models and AGI architectures rest, in their essence, on formal logic, Bayesian probability, and inductive pattern matching. Yet the foundation of human thought and of scientific leaps is dialectical logic.

Parameter	Formal / Statistical Logic (LLM)	Materialist Dialectical Logic (Human Cognition)
Basic Mechanism	Statistical distribution in the dataset and probability calculation	Grasping and negating contradictions (Aufhebung)
Limit and Scope	Remains within the limits of past data (Induction)	Creates paradigm shifts and radical epistemological ruptures
Criterion of Reality	Formal consistency and compatibility with data	Transformative Practice (Praxis) and Historical Objectivity
Error and Contradiction	Contradiction is eliminated as an error (hallucination)	Contradiction is grasped as the internal motor of development

The Impossibility of the Dialectical Leap:

An artificial intelligence model can process trillions of data belonging to Newtonian physics and simulate Kepler’s laws to perfection. Yet it cannot realize an Einsteinian epistemological rupture that senses the internal contradictions of existing physics and radically redefines the concepts of space and time.

For dialectical thought requires philosophically embracing contradiction and constructing new concepts through the negation of the negation. Algorithmic systems, by contrast, are formal structures programmed to avoid contradiction and to suppress it as statistical deviation.

The Epistemological Cover for Cognitive Dispossession

The positivist promises of Demis Hassabis and the techno-feudal elites that "AGI is coming in 2 years" rest epistemologically on the rejection of materialist dialectics and praxis. This approach, which severs intelligence from social labor, the body, and history and imprisons it in chips, is nothing but an ideological cover for those who wish to dispossess humanity’s collective intelligence and accumulation and place them at the command of capital.

The Class Map of the Western Cognitive Bloc: Techno-Feudalism and State-Monopoly Capitalism

Silicon Valley’s promises to "liberate humanity," "democratize artificial intelligence," or "create a society of abundance with AGI" are contemporary ideological curtains produced by the monopoly bourgeoisie to anesthetize the masses. Behind this structure stands State-Monopoly Cognitive Capital, which dispossesses humanity’s common historical accumulation, crushes living labor between the gears of dead labor, and is directly articulated to American imperialism’s war machine.

The thesis that V. I. Lenin advanced in Imperialism, the Highest Stage of Capitalism—"the monopolization of capital, the birth of finance capital, and the complete fusion of the monopoly bourgeoisie with the state apparatus"—has today reached a new dimension through cognitive capital and data infrastructure.

The Cognitive Monopolization of Capital and the Theft of the General Intellect

Karl Marx, in the famous "Fragment on Machines" section of the Grundrisse, notes that with the development of capitalism, social knowledge, science, and technique would be transformed directly into a productive force, and that capital would absorb this structure into itself as the General Intellect (General Intelligence / Social Knowledge-Power).

Today the structure of Silicon Valley that we call the Western Cognitive Bloc is precisely the expropriation of this General Intellect:

Cognitive Dispossession: The accumulation of language, art, science, philosophy, and code that humanity produced over thousands of years—that is, collective living labor—has been transferred into enormous data centers without any payment whatsoever.
The Domination of Dead Labor: This accumulation of knowledge processed in data centers is transformed, in Marx’s phrase, into dead labor (constant capital). After the models are trained, this dead labor is brought forth against living cognitive labor (the programmer, translator, academic, artist, engineer) as an enormous power of capital that threatens to render it unemployed or to drive down its wages.
Techno-Feudal Rentierism: Cognitive monopolists have appropriated not so much the means of production as "cloud infrastructure" and "model access." This structure, which collects a cognitive rent (cognitive rent) at every step over the value created, resembles a digital feudalism more than a classical capitalist market.

The Class Profile of the Techno-Feudalists and Military-Intelligence Integration

The actors who govern the Western Cognitive Bloc are not isolated "entrepreneurs"; they are organs of monopoly capital, the American security state, and the military-industrial complex.

Sam Altman & Greg Brockman (OpenAI / Microsoft): Behind the non-profit foundation illusion founded with the promise of "developing safe AGI for the benefit of all humanity" stands the technocratic bourgeoisie fused with Microsoft capital. OpenAI has removed the bans on military and warfare use from its usage policies and, by including former NSA Director General Paul M. Nakasone on its board of directors, has been directly integrated into the US national security architecture.
Demis Hassabis & Sundar Pichai (Google / DeepMind / Alphabet): Under the cover of positivist scientism, they hold the monopoly on digital search and data collection. They are the principal executors of Project Maven, in which drone imagery is analyzed with artificial intelligence for the US Department of Defense (Pentagon), and of billion-dollar cloud contracts with the Pentagon/CIA (JWCC).
Dario & Daniela Amodei (Anthropic / Claude): With "Constitutional AI" and ethical discourses, they represent the liberal technocratic bourgeoisie. Anthropic, which stands on Amazon and Google funding, has integrated Claude models with Palantir for use in classified intelligence and defense networks (AWS Secret Cloud).
Elon Musk (xAI / SpaceX / Tesla): He is the representative of reactionary, far-right oligarchic capital. SpaceX’s Starshield satellites and xAI’s infrastructure are directly linked to the US Department of Defense’s and the military’s special artificial intelligence networks (GenAI.mil).
Peter Thiel & Alex Karp (Palantir Technologies): They are the military-intelligence backbone of the techno-feudal ecosystem. Palantir was founded with the direct support of the CIA investment fund In-Q-Tel. It takes the models of OpenAI, Anthropic, Meta, and Google and transforms them into systems of target identification, mass surveillance, and autonomous destruction for the Pentagon, CIA, NSA, and the Israeli Defense Forces (IDF).

The Class and Operational Map of the Western Cognitive Bloc

Company / Actor	Source of Capital and Alliance	Ideological Mask	Class Function and Essence	Military and Intelligence Integration
OpenAI (Altman, Brockman)	Microsoft ($13B+), Venture Capitals	"AGI for the Benefit of Humanity and Open Science"	Dispossession of cognitive labor and data; monopoly capital accumulation.	Former NSA Director (Gen. Nakasone) on the board. Integration with US Cyber Command.
Google / DeepMind (Hassabis, Pichai)	Alphabet Monopoly, Global Advertising and Data Rent	"Positivist Science and Rational Problem-Solving"	Holding cognitive property; automating digital labor processes.	Project Maven, Pentagon Cloud Contracts (JWCC), CIA/NSA data analytics.
Anthropic (Amodei Siblings)	Amazon ($4B), Google ($2B)	"Ethical, Safe, and Constitutional Artificial Intelligence"	Liberal bourgeois legitimization; capital’s risk control.	Operational use on Pentagon IL6 (highest classification) networks through partnership with Palantir and AWS.
xAI / SpaceX (Elon Musk)	Far-Right Oligarchic Capital, State Subsidies	"Absolute Freedom of Expression and Truth AI"	An instrument of the reactionary bourgeoisie for suppressing labor and deregulation.	Starshield military satellite network, Pentagon GenAI.mil infrastructure, and live battlefield data flow.
Palantir (Thiel, Karp)	CIA Fund (In-Q-Tel), Pentagon Contracts	"Protecting Western Civilization and Democracy"	The military-intelligence backbone that transforms cognitive models into weapons.	Live targeting and surveillance systems for the Pentagon, CIA, NSA, FBI, and the Israeli Defense Forces (IDF).

The Military-Cognitive Complex: The Transformation of Models into Weapons of War

While developing cognitive technologies, capital aims not only to drive down the wages of the working class but also to perfect the instruments of imperialist domination. The destination of the Western Cognitive Bloc is the Military-Cognitive Complex.

The operation of this mechanism unfolds in three fundamental operational stages:

Stage 1 - The Plunder of Data: The digital footprints of the working class, intellectuals, and the masses are absorbed as unpaid raw material.
Stage 2 - Transformation into Models: OpenAI, Anthropic, or Google process this data with enormous computing power (compute) and transform it into language and decision-making models.
Stage 3 - Weaponization and Integration: Integrators such as Palantir, AWS Military Cloud, and Azure Military Cloud take these models and combine them with live data streams from satellites, drones, and listening stations. The result: autonomous destruction and surveillance mechanisms that turn people into targets within seconds in Gaza or in any oppressed geography of the world.

The Eastern Alternative: The Chinese State, Revisionism, and the "DeepSeek" Reality

In recent global artificial intelligence debates, especially in left-liberal, neo-Kautskyite, and techno-utopian circles, a mystical legend has been constructed around the "cheap, efficient, and open-source Chinese artificial intelligence rising from the East." While DeepSeek’s algorithmic optimizations shake Silicon Valley’s trillion-dollar balloon capital expenditures (CapEx), some circles hail this situation as the "collapse of capitalism" or the "victory of public technology."

Yet a dialectical materialist and epistemological analysis shatters this rosy picture to pieces. The reality is this: China-centered artificial intelligence models are not the victory of socialism or of a social form of production; they are the technical response of the State Capitalism constructed by the post-Mao bureaucracy of the Communist Party of China (CPC) to the global crisis of capital accumulation, and a new front in the rivalry between imperialist blocs.

Epistemological Analysis: Lenin, Ilyenkov, and Bureaucratic Dogmatism

In the Philosophical Notebooks, while annotating Hegel’s Science of Logic, V. I. Lenin defines the essence of dialectical epistemology as follows:

"Dialectics is the theory of knowledge of (Hegel and of) Marxism... It is the ability to grasp the contradictions within the object itself, its own movement, its own life."

When this thesis of Lenin’s is taken together with the theory of reflection in Materialism and Empirio-Criticism, the epistemological limits of China’s state-capitalist artificial intelligence models become clear:

Formal Logic vs. Dialectical Logic (The Examination of Kopnin and Kedrov): As the Soviet philosophers Pavel Kopnin and B. M. Kedrov showed, formal logic treats objects within fixed, external, and contradiction-free relations. Chinese models such as DeepSeek, Qwen, or Baidu rest on exactly the same formal-logical and statistical base as their Western rivals. What is done is merely the compression of formal logical reasoning processes (Chain-of-Thought) onto cheaper chips and smaller parameters. There is no philosophical leap there, only a technical compression.
Ilyenkov and "The Censorship of the Ideal": In The Dialectic of the Ideal, Evald Ilyenkov shows that the "Ideal" is the social-cultural form that is materialized as a result of human social praxis. The Chinese state has placed the "Ideal" under bureaucratic control in the artificial intelligence models it develops. In Chinese models, the dataset is pruned within the ideological limits approved by the CPC bureaucracy. Contradiction, instead of being counted as the motor of development, is erased from the algorithm as a "systemic error" in the name of bureaucratic stability.

The Political Economy of DeepSeek: The Tendency of the Rate of Profit to Fall and Constant Capital

DeepSeek’s (R1 and V3) reduction of training costs to 1% of those of the Silicon Valley giants (approximately 6 million dollars) was presented in the bourgeois media as a "miracle." Yet dialectical materialism shows us the economic essence of the process.

Karl Marx explains the law of the tendency of the rate of profit to fall in Volume 3 of Capital. According to this, the Rate of Profit is obtained by relating the Surplus Value produced to the sum of Constant Capital (buildings, machines, raw materials, computing power) and Variable Capital (wages paid to the worker).

The Silicon Valley giants (OpenAI, Google, Meta), by investing hundreds of billions of dollars in enormous chip infrastructures, data centers, and power plants in the artificial intelligence race, have inflated the constant capital item to an extreme degree. This enormous growth in constant capital, even if the mass of surplus value produced increases, has triggered the law of the tendency of the rate of profit to fall on a global scale.

What DeepSeek has done is to radically reduce the requirement for constant capital through innovations at the algorithmic and architectural level (MoE, MLA, FP8 precision):

The Cheapening of Constant Capital: In Chapter 14 of Volume 3 of Capital, Marx counts the cheapening of the elements of constant capital among the foremost "Counteracting Influences to the Fall in the Rate of Profit." DeepSeek, by optimizing the algorithm, has lowered the electricity and GPU cost required to produce a token; thus it has opened space for global capital’s struggle to preserve the rate of profit.
Price Undercutting Among Capitals (Dumping): DeepSeek, by presenting this optimization "as cheap/open source to global developers," has undermined the high profit margins extracted by American monopoly companies. This is not socialist internationalism; it is the strategy of Chinese state capitalism to penetrate the market, crush rivals with cost pressure, and set standards.

The Class and Operational Matrix of China’s Cognitive Giants

Actor / Company	Ownership and Capital Structure	State and Party (CPC) Integration	Epistemological and Ideological Function	Global Capital and Military Strategy
DeepSeek (High-Flyer Capital)	Quantitative Hedge Fund (High-Flyer) capital; a direct derivative of finance capital.	Indirect/direct strategic alliance with CPC technocracy and state funds.	Algorithmic efficiency; positivism that reduces dialectics to formal calculation.	Bypassing US chip embargos; undermining Western cognitive rent with cheap models.
Alibaba (Qwen / Tongyi)	Private-State Hybrid Monopoly Capital.	CPC Committee within the company; integrated with state logistics and data infrastructure.	Market rationalism that automates consumption behaviors and trade networks.	Creating infrastructural dependency with open models in the Global South market.
Baidu (ERNIE Bot)	Cognitive Digital Capital Monopoly.	Direct administrative oversight relation with the Cyberspace Administration of China (CAC).	An ideological filtering mechanism that produces and preserves party-state discourse.	Protecting the domestic market from foreign capital; sustaining the national cognitive market monopoly.
Huawei (Ascend / Pangu)	Strategic State Company.	Direct ties to the People’s Liberation Army (PLA) and state intelligence.	Nationalist-Statist infrastructural hardware consciousness.	Alternative chip/cloud hegemony on the Asia-Africa axis against Nvidia’s hardware monopoly.

The Class Illusion of the "Open-Weights" Strategy

The publication of DeepSeek’s and Alibaba’s models as "open-weights" is hailed in left circles as a "communizing sharing." Yet dialectical materialism looks not at the appearance of facts but at the relations of production behind them. The open-source strategy, at capitalism’s developed stage, means the plunder of unpaid R&D labor:

Free Cognitive Labor Extraction: Thousands of independent programmers and engineers worldwide fix the models’ errors and write new code. Chinese state-capitalist companies absorb this enormous collective living labor by paying zero wages and incorporate it into their own ecosystems.
Dependency Through Software Standards (Ecosystem Lock-in): The open presentation of models renders developers worldwide dependent on China-origin code libraries and, over time, on Huawei’s Ascend chip architecture. This situation is the digital and software form of "the export of capital" in Lenin’s theory of imperialism.

Petty-Bourgeois Romanticism: The Illusion of "Open Source" and "Social AI"

The most insidious and seductive maneuvers of bourgeois ideology are not always launched from directly within the fortress of capital; they often seep in from the ranks of petty-bourgeois utopianism, of the academy carrying a "left" label, and of techno-romantic circles.

This search appears before us today in the form of "Open-Source Liberation," "Social AI / AI Buddy Projects," "the People’s Artificial Intelligence," or "Decentralized Cognitive Communities" (Hugging Face, EleutherAI, Mozilla AI, etc.). Yet this discourse is nothing other than a renewed Proudhonism of the twenty-first century and an illusion of Aleksandr Bogdanov-style Proletkult.

Epistemological Roots: Lenin, Bogdanovite Proletkult, and the Sophistry of "Pure Science"

In Materialism and Empirio-Criticism, V. I. Lenin also sharply criticized Aleksandr Bogdanov, the representative of the "left" deviation within the Party. Bogdanov, setting dialectical materialism aside, constructed a positivist system he called the "organization of experience" and argued that he could build an isolated "pure proletarian culture and science" within capitalist society.

The Illusion of Science Above Classes: Bogdanovite idealism has returned today in the guise of "independent scientists’ open-source communities." To present science and software as a "sterile" domain severed from class struggle, from the ownership of the means of production, and from the state apparatus is to surrender directly to bourgeois idealism.
The Class Character of Data: Communities that claim to develop "Social AI" again use the data dump that capitalist society has left on the internet to train their models. While the data itself bears the traces of capitalist relations of production, to believe that an "independent, ethical, and social consciousness" will emerge from this data is an epistemological ignorance.
Humanism Severed from Praxis: As Ilyenkov showed, the human "Ideal" and consciousness are shaped not by code written in sterile laboratories or on Discord servers, but within social-historical praxis. A "Social AI Buddy" severed from class struggle cannot go beyond the bourgeoisie’s digital opium.

Proudhonism 2.0 and Idealist Craftsmanship

Karl Marx, while shattering Pierre-Joseph Proudhon’s petty-bourgeois economic understanding in The Poverty of Philosophy (1847), exposed Proudhon’s dream of eliminating capitalism’s "bad sides" (monopoly, misery) while preserving its "good sides" (craft, free struggle). Today’s champions of "Open Source" are precisely Proudhon’s Heirs in the Cognitive Age:

The Craftsman’s Illusion: They believe that openly sharing the model’s parameter weights (open-weights) liberates technology from monopoly capital. This resembles leaving the factory building to the monopolist in a capitalist world and rejoicing that "we have socialized the machine’s user manual."
The Monopoly of Material Infrastructure: The objective elements required to train an artificial intelligence model—hundreds of thousands of GPUs, enormous data centers, power plants, chip foundries—are entirely in the ownership of monopoly capital and imperialist states. The openness of the manual does not change the ownership of the factory.

Capital’s Cognitive Plunder: Free R&D and the Transfer of Surplus Labor

The open-source movement, while imagining itself "anti-capitalist," turns into monopoly capital’s unpaid R&D department. Meta’s publishing of Llama models as "open," or Amazon’s funding of the Hugging Face platform, is a cold-blooded capital strategy.

When we look through Marx’s law of value, Total Cognitive Capital Accumulation is composed of the sum of these three components:

Constant Capital: Data centers, chips, and electricity infrastructure.
Paid Variable Capital: The wages of staff engineers within the company.
Unpaid Collective Living Labor: The cognitive surplus labor offered without compensation by open-source communities and independent programmers.

Big Tech companies incorporate this unpaid cognitive labor component into themselves at zero cost. While radically lowering their internal R&D costs, they appropriate for free the idealist living labor of thousands of young minds. The next day these codes are closed by Microsoft (Azure) or Amazon (AWS) and transformed into a commercial service sold by the hour for dollars.

Comparative Class and Epistemological Matrix

Plane of Analysis	Petty-Bourgeois Open Source / "Social AI" Illusion	Dialectical Materialist and Class Reality
Conception of Property	The openness of parameter weights and code is sufficient for liberation.	As long as the material infrastructure (GPU, chip, power plant, data center) remains in monopoly hands, code openness is an ideological cover.
Subject and Class	"Scientists, communities, and ethical developers independent and above classes."	Cognitive proletarians producing unpaid R&D and surplus labor for monopoly capital.
Epistemological Foundation	Bogdanovite Proletkult and the Machist conception of "pure experience / neutral data."	Materialist Dialectics; data is class-bound; consciousness takes shape only through social praxis.
Relation to Capital	Living in decentralized "social" oases alternative to Big Tech.	An indirect relation of exploitation in which the value produced is absorbed by Big Tech at zero cost.
Political Strategy	Building "ignore" or "parallel technology" by founding communities within the system.	Revolutionary political struggle for power that will abolish the private ownership of the means of production.

Global Cognitive Capital Rivalry and the Revolutionary Tasks of the Working Class

Comrades, the picture standing before us is not reducible solely to the greed of American techno-feudalists, nor to the technical leaps of Chinese state capitalism, nor to the idealist good intentions of petty-bourgeois developers. The phenomenon before us is the struggle by imperialist blocs, in capitalism’s historical stage of crisis, to plunder and repartition cognitive infrastructure and humanity’s common intelligence (General Intellect).

Viewed within the framework of Lenin’s analysis of finance capital, the rivalry of cognitive capital is conducted along three fundamental axes:

The Monopoly of Material Infrastructure and Hardware: Chip design (Nvidia), lithography machines (ASML), and foundries (TSMC), which form the backbone of global computing power, are under the control of the Anglo-American imperialist bloc. Chinese State Capitalism (Huawei, SMIC) is pursuing the path of algorithmically lowering the cost of constant capital in order to break this hardware monopoly.
Data Plunder and Dispossession: All of humanity’s historical and contemporary cognitive production is extracted as unpaid raw material; it is commercialized and brought before society as a mechanism of cognitive rent (cognitive rent).
Military and Biopolitical Domination: While the American Cognitive Bloc (Palantir, OpenAI, Anthropic, Pentagon) transforms this technology into autonomous destruction systems and global surveillance networks, Chinese State Capitalism (Baidu, CAC, PLA) constructs a biopolitical apparatus of control that dampens social opposition.

Comparative Matrix of the Blocs of Global Cognitive Capital

Bloc / Line	Economic and Ownership Structure	Epistemological Foundation	Class Function and Essence	Revolutionary Critique and the Working Class’s Stance
Western Cognitive Bloc (US Techno-Feudalism)	Monopoly Cognitive Capital, Wall Street Funds, Pentagon Contracts	Machist Data-Criticism, Bayesian Statistical Positivism	Dispossessing living labor, collecting cognitive rent, automating war.	The principal center of the imperialist war machine and techno-feudal dispossession; uncompromising struggle and the goal of socialization.
Eastern Cognitive Bloc (Chinese State Capitalism)	Bureaucratic State Capitalism, CPC-Capital Alliance, Quantitative Hedge Funds	Statistical Positivism + Bureaucratic Dogmatic Filtering	Lowering the cost of constant capital, seizing market share, biopolitically controlling social opposition.	Rejection of the revisionist state-capitalist illusion; an independent class line against capital accumulation in socialist guise.
Petty-Bourgeois Line (Open Source / "Social AI")	Utopian Craftsmanship, Unpaid Collective Cognitive Labor Pool	Bogdanovite Proletkult, Idealism of "Pure Science/Data" Above Classes	Being a source of free R&D and surplus labor for Big Tech and State Giants.	Exposure of Proudhonist and Bogdanovite dreams; preventing youth energy from becoming free R&D for capital.
Horizon of the Revolutionary Proletariat (Marx-Lenin-Ilyenkov Line)	Social/Collective Ownership of Cognitive Infrastructure and the Means of Production	Materialist Dialectics, Theory of Reflection, Theory of Praxis	Shortening the working day, ending alienation, securing the all-round development of the human being.	Without falling into hostility to technology (Luddism), socializing all infrastructure through the revolutionary struggle for power.

The Class Stance of the Working Class and Revolutionary Tasks

For a dialectical materialist philosopher, contemplation does not remain limited to interpreting the world; it must point to the historical subject and the revolutionary practice that will change it. The clear principled positions we must adopt in the face of artificial intelligence and cognitive capital are these:

The Absolute Rejection of Techno-Luddism (Hostility to Technology)

Karl Marx, while criticizing machine-breaking (Luddism) in Volume 1 of Capital, notes that it took time for workers to learn the difference between the machine itself and the capitalist use of the machine. Artificial intelligence is humanity’s accumulated dead labor. The problem is not algorithms or chips; the problem is that these instruments are in the ownership of the bourgeoisie. The revolutionary stance is not to be hostile to technology but to liberate it from the yoke of capital.

Exposure of Revisionist and Utopian "Third Ways"

Neo-Kautskyite deviations that present Chinese State Capitalism as an "anti-imperialist or socialist alternative" must be fought with resolve. DeepSeek or Qwen were designed not for the emancipation of the Chinese working class but to protect the profit rates of Chinese state capitalism. Credit must not be given to open-source romanticism or to utopias of "parallel technology" imagined to be built on Discord servers. No code can be free until property relations are overthrown.

The Strategy of Socializing Cognitive Infrastructure

The only path for artificial intelligence to become a liberating force is the dispossession of the means of production. We must pass from the capital order that seizes data centers, GPUs, and social labor to Social Planning.

All data centers, cloud infrastructures, chip production facilities, and artificial intelligence models must be torn from the hands of private companies and bourgeois states. Under the dictatorship of the proletariat and social planning, this cognitive power must be used not to render the working class unemployed or to surveil it, but to reduce compulsory working hours (the daily working time to 3–4 hours), to rationalize production planning, and to remove the obstacles before the human being’s mental/bodily development.

Historical Dialectics, Cognitive Alienation, and the Proletariat’s Ultimate Horizon

Comrades! Demis Hassabis’s promises that "AGI is coming within 2 years," Sam Altman’s rosy fairy tales of abundance, and Silicon Valley’s techno-feudal cheerleading are, philosophically speaking, nothing but subjective idealist instances of digital commodity fetishism striving to keep stock-market speculation alive. While these charlatanries cover over the deepening crises of finance capital, in the back rooms they construct, through the Palantir and Pentagon networks, a military-cognitive mechanism of destruction that turns humanity into autonomous targets and surveils the masses.

On the other side, while Chinese state capitalism, in this global race, seeks to lower the cost of constant capital against the tendency of the rate of profit to fall and to seize market share, petty-bourgeois open-source romantics carry unpaid R&D labor to the enormous gears of monopoly capital in GitHub repositories and Discord servers.

Dialectical materialist analysis shows us the naked truth: Artificial intelligence is not a mystical consciousness descending from the heavens or a miraculous object substituting for the biological brain. Artificial intelligence is the form in which the collective knowledge humanity has produced throughout the historical process—that is, "living labor"—has been dispossessed by capital and crystallized in data centers.

In Marx’s expression, it is an enormous mass of dead labor that stands upright by sucking the blood of the living worker.

Yet in accordance with the most fundamental law of dialectics, this most developed productive force that capital has created with its own hands rests at the same time on a historical contradiction that prepares its own end:

The Ultimate Blockage of the Law of Value: In the capitalist mode of production, the sole source of value and surplus value is human labor. To the extent that capital, in its greed for profit, drives living human labor out of production and puts artificial intelligence and automation in its place, it ironically dries up the sole source of its own existence and profit. Full automation means the absolute collapse of capitalist market relations and of the value form.
The Summit of Historical Alienation (Entfremdung): The social intelligence (General Intellect) that the human being created with their own hands and mind has, under capital ownership, become alienated from the human being and transformed into a hostile power that threatens them with unemployment, precariousness, cognitive impoverishment, and autonomous weapons.

This historical knot cannot be resolved by bourgeois reforms, sermons on "ethical artificial intelligence," or petty-bourgeois craftsmanship. The solution is hidden in a dialectical leap—that is, in class struggle.

In Whose Service Will Artificial Intelligence Be?

The historical question standing before us is not a techno-utopian fantasy but directly a question of power: Will artificial intelligence be a weapon of domination and an instrument of surveillance for the techno-feudalists who enslave humanity; or will it be the infrastructure of humanity’s liberation from the vise of compulsory labor?

The answer of the revolutionary proletariat is clear: We are not technology-hostile Luddites! We are the historical subject standing against capital’s expropriation of humanity’s common mental heritage.

The Proletariat’s Revolutionary Cognitive Hour:

Expropriation of the Expropriators: All data centers, chip foundries, cloud infrastructures, and algorithmic models must be torn from the hands of techno-feudal companies and bourgeois states and socialized.

Transition to the Realm of Freedom (Reich der Freiheit): This technology, which in the bourgeoisie’s hands is a threat to render the masses unemployed, will under the dictatorship of the proletariat and social planning be used to radically reduce the daily compulsory working time to 3–4 hours.

The All-Round Emancipation of the Human Being: With the takeover by algorithms of alienating and repetitive mental/bodily tasks, the human being will become an all-round being who engages directly with science, art, philosophy, and their own social self-administration.

Young comrades armed philosophically with materialist dialectics, economically with Marx’s law of value, and politically with the Leninist revolutionary line!

Our promise is not the bourgeoisie’s false digital heavens or avatar simulations. Our promise is a concrete, objective, and revolutionary praxis that will put an end to the exploitation of human by human!

We will socialize those enormous data centers and chip factories in the hands of capital and the imperialist war machines, and we will liberate the collective human mind from its chains!

Down with Techno-Feudalism, Down with Imperialism and State Capitalism!

Down with Capital’s Cognitive Yoke!

Long Live Humanity’s Struggle for Cognitive and Social Emancipation!