Being, Alienation, and Production in the Age of Generative Artificial Intelligence

1.1. Introduction

Since the first public release of OpenAI’s GPT-3 model in 2022, large language models (LLMs), among other generative artificial intelligence technologies have become increasingly ubiquitous across classrooms, labor, industry, and digital spaces. Researchers and boosters alike have characterized the growth of AI applications as equating to “Industrial Revolution 4.0,” responsible for “turning the whole world upside down, replacing humans in more than several sectors by establishing new ways of doing things (in more optimizable and efficient routines)” (Rashid and Kausik 2024, p. 1). While the rapid sophistication of LLMs and other generative AI tools in recent years is undeniable, the similarly rapid adoption of this technology across nearly all facets of contemporary life poses numerous questions as to whether such novel means of optimizing and accelerating will prove truly beneficial in the long term.

Some commentators assert the continued growth of AI tools may lead humanity through a period of “liberatory alienation” (Sidorkin 2024, p. 1417). This view proposes that, through the mass offloading of menial, manual, and cognitive labor onto intelligent machines, humans will be “liberated to focus on uniquely human qualities such as creativity, agency, and the capacity for joy” (p. 1417). Having achieved the Homo Liberatus, humans will be able to “transcend” the constraints of their physical environment, and “step toward a more liberated, intellectually engaged human existence” (p. 1422). While this AI-powered utopia sounds pleasant, the reality of AI-driven cognitive offloading appears far less ideal. A recent MIT study on the cognitive impact of generative AI use revealed a 55-percent decrease in neural activation in the brains of those tasked to write a series of essays with the assistance of LLMs over a four month period (Kosmyna et al. 2025, p. 136).

Researchers divided participants into three groups: the LLM group, the search engine group, and the brain-only group (2025, p. 2).1 In broad strokes, these groupings determined what possible series of operations might be employed by participants in completing the assigned essay. While the use of ChatGPT remained optional for those in the LLM group, participants all engaged with the tool at one point or another over the course of the study, often beginning by consulting ChatGPT for pointers on linguistic content or essay structure (2025, pp. 31, 34, 35). In these cases, the offloading of cognitive labor onto large language models superseded the prior process in which subjects would have thought, planned, and worked through the essay. This new process — in which consultation with LLMs serves as the first step — is defined by an active cognitive offload of ranging significance and scale. While LLM group participants initially demonstrated greater efficiency and productivity, they would perform far behind the brain-only group in each category of evaluation over the course of the study.2

The impact of this differentiation in process was reflected in researchers’ analysis of EEG data, which displayed distinct impacts on cognitive structure between each group. Participants were found to have developed “clearly different neural connectivity patterns, reflecting divergent cognitive strategies… with implications for executive function, semantic processing, and attention regulation” (2025, p. 135). These structural distinctions were exhibited behaviorally in the written content, memory recall, and senses of ownership of the compositions. In the case of the LLM-group, researchers found a striking statistical homogeneity in both essay content and structure across sessions (p. 136). Additionally, LLM-group participants remained hesitant to take full ownership over submitted essays, and were largely unable to accurately recall or quote the contents of essays written “just minutes prior” with 83-percent of LLM-groupers failing to do so in the first session, and 33-percent by the third session (pp. 2, 133, 137). In contrast, brain-only and search engine group participants demonstrated “near-perfect quoting ability” by the second session and exhibited greater variability in content, with the brain-only group demonstrating the most variability (pp. 137, 133).

Across the first three sessions, those using LLMs became increasingly reliant on the tools in their approach to writing, suggesting a possible erosion in critical thinking and problem-solving capacity due to a lack of cognitive engagement (2025, p. 112). This pattern reflects the possible accumulation of what researchers termed cognitive debt, defined as “a condition in which repeated reliance on external systems like LLMs replaces the effortful cognitive processes required for independent thinking” (p. 141). When one is fully reliant upon an LLM for the generation of writing, the building of understanding, or the attempted production of creative thought, that individual is robbed of epistemic agency and the critical neural activation involved in problem-solving. Lower neural activation entails lower connectivity over time, eroding capacity for independent, critical, and creative thought in its wake (p. 112).

Importantly, this investigation was limited to college students and graduates, though recent surveys show that LLM use is widespread, with 69-percent of students and 43-percent of workers reporting semi-regular use of the tools (Gillespie et al. 2025, pp. 68, 90).3 The growing ubiquity of these tools — coupled with their demonstrated impact on cognition and behavior — begs the question of what a widespread accrual of cognitive debt might spell in the long term, especially within the developing brains of children.

The striking preliminary results of the MIT study further underline the importance of self-direction in mental production and activity more generally. Meaningful self-direction in activity promotes meaningful engagement of the brain, fostering greater neural interconnectivity and a higher sense of embodiment within one’s actions. This issue of ownership and embodiment over one’s work, while at a small scale in the MIT study, offers a unique opportunity to directly interrogate the manner in which these generative technologies are reshaping our relations to production at the social and cognitive levels. Furthermore, the incisive implications of this study provide the grounds to theorize how such tools may be alienating us in distinctly harmful, non-liberatory manners. As greater sums of cognitive labor and mental production become offloaded to AI technologies, will the accrual of cognitive debt estrange us from our neural capacities, and further alienate us from meaningful productive engagement?

Much of the production humans engage in today is mediated through various labor divisions, organizational structures, and is partially, if not wholly reliant upon contemporary digital technologies. As a result, the form and content of production, alongside the degree of self-direction within it, remain incredibly variable. In some cases, labor can appear fully disembodied and externally directed. Labor under these conditions often engenders a sense of disconnection with one’s work — a state commonly referred to as alienation within the Marxist canon (Haga 2020, p. 801).

With this paper I resurrect and elaborate this concept to approach generative AI technologies as novel mechanisms of alienation. I argue that the off-loading of the baseline cognitive capacities involved in critical thought, concretized into “intelligent” machinery, constitutes an ongoing estrangement of the individual from her being — or her ability to produce, engage in, and externalize conscious, critical thought. This pattern, I demonstrate, emerges not solely from the development of intelligent technologies and algorithms in themselves, but as a manifestation of the underlying market logic shaping the production and deployment of AI technologies. This pattern in turn limits the productive development of both AI technologies and cognitive engagement, and is representative of a tendency towards the generalization of production into forms most adequate to capital.

The course of industrialization in the 19th century was driven by a logic of mass production aimed at minimizing cost and maximizing profit. Mechanization allowed for significant cuts to labor cost, while simultaneously enabling new leaps and bounds in generalized productive force and capital generation. The development of generative technologies today follows a similar tendency. Like the machines of Marx, generative AI technologies lie outside the human as alien bodies of production. These tools are frequently advertised as inherently superior to the mental production of humans, with boosters prophesying the elimination of human jobs across any number of fields (Rashid and Kausik 2024, p. 1). Between the growing role generative AI technologies occupy in mediating information dispersal and shaking up forms of contemporary labor; their active disembodiment of mental production in and beyond labor, as well as their deleterious impacts upon cognitive capacity, these tools appear to be novel forces reshaping our brains, social relations, and our relations to production more generally. As such, extending critical theorization towards existing discussions in AI studies situated within the implications of emerging quantitative research, we may establish a material basis to approach these technologies in a framework of an extended Marxian alienation.

1.2. A Brief on Alienation

Marx first describes alienation in his Economic and Philosophic Manuscripts (EPM) as the process through which workers come to occupy an increasingly abstract position within the scope of production. The result is a state in which the worker stands apart from (1) the product of her labor, (2) the activity of production (or species-activity), (3) her fellow human, and (4) her very being (or species-being) as human (Marx 1964, pp. 122-124, 127, 129). Human species-activity refers to how the human utilizes the resources in her surrounding environment and community to survive. The human engages in its species-activity as would any other organism, utilizing the organic and inorganic means at her disposal to achieve and maintain subsistence. What differentiates human activity in-nature from that of animals, for Marx, is the manner in which productive human activity meets, and extends beyond bare necessities. The human, in community with other humans, engages in conscious, self-directed, material, intellectual, and social production towards achieving security. Importantly, she continues to do so once she has reached this state. It is the confluence of the creative and conscious expression within human species-activity which gives way to the species-being (1964, pp. 126-127). In this account, the alienated life is no longer defined by self-direction, rather, it is directed towards wage labor, a mere means of subsistence largely determined by external, alien forces (p. 127).

Alienation, for Marx, is an external material process with material, social, and psychological impacts on workers. While the sense of distance and disaffection from one’s labor remained a clear concern motivating Marx’s investigation, his primary worry across the EPM lay within the material processes and economic relations that yield an alienating productive arrangement. The manner in which production is organized within a given society provides the basis upon which that society produces and reproduces its social relations. At the time of Marx’s writing, industrial machinery was actively transforming the form and function of production across England. Later variations of Marx’s alienation center the growth of automation and machinery as significant catalysts driving the rupture between workers and production, with machinery either appropriating or dictating the activities of individual laborers (Marx 1993, pp. 692-695).

Any investigation into novel productive forces must prioritize close analysis of the underlying logic of the system which produces certain productive arrangements. As we will see, approaching generative AI tools within a wider framework of alienation requires significant reconsideration towards the relationship between production and technology more generally. To build a supplementary understanding of technology and machinery in general, I recast alienation through the development of Marx’s thought in the Grundrisse. Marx excels in the Grundrisse at outlining the specific relationships between labor, machinery, and the process through which machinery comes to appropriate labor away from workers. However, this account remains limited insofar as Marx’s focus is on the “general aspect” of machinery, namely, how “the means of labour…loses its direct form, becomes fixed…and confronts the worker physically as capital” (Marx 1993, p. 695). As such, I build upon a more general technical understanding through the work of a Young Hegelian peer of Marx, Ernst Kapp, alongside the contemporary theory of Raymond Ruyer, both of whom provide valuable consideration to forms of production which, in Kapp’s case, decenters the logic of capital, and in Ruyer’s, decenters the human as its primary source.

1.3. Structure and Overview

Section two (2.1-2.3), provides an overview of Marx’s alienation, with particular attention given to its earlier formulation and its critical reception by 20th and 21st century scholars. I assert the importance of alienation’s evolution in the Grundrisse, specifically examining the manner in which labor activity comes to occupy fixed capital formations.

In section three (3.1-3.2) I argue that Kapp’s dialectical view of tools as externalizations of the body’s organs offers a complementary account of unalienated productive relations. I introduce the works of Raymond Ruyer in the fourth section (4.1-4.2). I mobilize his epistemic concerns alongside those of Marx to diagnose the limitations of Kapp’s approach. I argue Ruyer and Marx are largely aligned in their concerns towards step-by-step causality, namely in how this causal interpretation is developed, and lends itself to flattened understandings of productive and historical development.

I further elaborate the logical implications of how production is conceptualized in the fifth section (5.1-5.4). I consider the deployment of alienation within Deleuze and Guattari’s Anti-Oedipus, a text which is largely indebted to Marx and Ruyer. I assert that the generalization of production advanced in Anti-Oedipus represents a conceptual continuation of Marx’s alienation in the Grundrisse, and draw upon each account to demonstrate how the combined logic of capital and causality has severely limited the developmental approach to language models, shaping how LLMs are produced and deployed. The result of this synthesis is then deployed within a heterodox Marxist theory of alienation. I apply this model to understand the novel “intelligence” technologies of generative AI in the sixth section (6.1-6.2), before concluding in the seventh section.4

2.1. Alienation: Origins and Motivation

Alienation remains a thematic constant across Marx’s work — especially so in his manuscripts. While the concept first appears, and is best known for its description in the EPM, his deployment of the concept evolves significantly across his later manuscripts. Despite the richness of this evolution, the form presented in the EPM has remained one of the most persistent and contentious concepts from Marx’s corpus, and has largely overshadowed its later development.

Throughout the EPM, we walk alongside Marx as he concocts his initial critique of English political economy. Marx begins from a point of reverence, drawing on the works of classical figures to outline the degradation of working conditions and the consolidation of capital into a “whole of society divide[d] into the two classes of property owners and propertyless workers” (p. 120). From this position, Marx sets out on his critique.

One of the largest issues Marx saw within classical political economy was its reliance on the presupposition of private property as a veritable fact of human existence. This reflected a larger pattern within the field, in which political economists abstracted economic formulas and extrapolated them as a priori laws reflecting human nature. Marx viewed such an approach as severely limited, unable to penetrate past its own frame of analysis. This restricted inquiry would birth a severely underdeveloped model of capitalism — built upon a fleeting notion of human nature with certain delimited external conditions impacting it. In Marx’s words: “what should be explained is assumed” (p. 120). Marx instead sought to flesh out the internal relations that defined the function of capitalism in all its complexity.

With his initial diagnosis of alienated labor, Marx aimed to characterize the process by which labor activity seemingly underwent continuous reformulation. Despite the staying power of the EPM’s alienation, a straightforward, concentrated definition is difficult to pin down when looking solely at this text. Bertell Ollman pins this difficulty on the lack of a concrete, external, “unalienated” standard with which to compare it to (1980, p. 132).5 Defining alienation from the EPM alone is further complicated by the manner in which Marx simultaneously discusses alienation as a process of appropriation and a subjective experience.

2.2. Marx’s Alienation: Essentialism and the Species-Being

While continuously referenced by sociologists and venerated among Marxist-Humanists, alienation has been subject to critical derision throughout the 20th century and into the present. Emil Øverseen divides these critiques into four primary areas — semantic overload, essentialism, paternalism, and conservatism (Øverseen 2022, p. 444).6 Øverseen argues that these issues have problematized the consistency and solidity of alienation, and largely grow out of interpretations which rely fully on Marx's earliest writings on the concept.

The driving contention across critical accounts is that Marxian alienation relies heavily, if not completely upon the notion of some essential, indelible human nature. Marx’s essentialism is frequently pinned to his use of species-being or species-essence. Marx’s species-being — inspired by Feuerbach — is defined by the species-activity of the human, or the process by which the human “makes the whole of nature into his inorganic body: (1) as a direct means of life; and equally (2) as the material object and instrument of his life activity” (1964, p. 126). Critics posit that through this appeal, Marx functionally adopts the existence of an essential character which the human has been separated, or made alien from. Among the most notable dissenters in this cohort was Louis Althusser, who rejected the concept on its humanist basis.

Althusser believed alienation and its essentialism was the direct result of the youthful Marx’s Hegelian streak, a tendency he believes the mature Marx abandoned. Althusser held the concept as a failure to fully escape the presupposition of political economist’s unified human, a misstep towards achieving a truly scientific Marxist analysis (Øverseen 2022, p. 443). Althusser asserted that Marxist thought, too, must abandon concern with alienation due to its unscientific essentialism. While Althusser’s suggested anti-humanism failed to pick up speed, the general thrust of his critique would.7 Since then, conversations surrounding alienation have largely grown distracted by attempts to fully reformulate it with the hopes of finally laying the essentialist critique to rest.

The more generalized, reformulated alienations often displace capital as a primary driving force. Rahel Jaeggi, following the essentialist critique, opts for a semi-existentialist approach. This alternative definition frames alienation as a “failure to identify with one’s own actions and desires and to take part in one’s own life” (Jaeggi 2014, p. 155). One grows alienated when systems or external forces interrupt one’s freedom and ability to appropriate, or act as the author of one’s own self-actualization (Jaeggi 2014, p. 39). Such a conceptualization is not wholly opposed to the earlier Marxist understanding of alienation; however, the tendency to separate alienation from capitalism, or Marx from alienation completely, mystifies the purpose, historical development, and analytic potential of the concept.

When approaching any piece composed by Marx — especially those published posthumously — one must take particular care to his often complex, even convoluted terminology. Ollman, quoting Vilfredo Pareto, laments that “Marx’s words are like bats: one can see in them both birds and mice” (Ollman 1980, p. 3). It becomes important then to distinguish Marx’s activity-centered species-being from the English political economists’ human nature, which relied heavily on inferred, natural social relations. Marx rails against the political economist’s human nature throughout the EPM, writing: “Let us not begin our explanation as does the economist, from a legendary primordial condition. Such a primordial condition does not explain anything; it merely removes the question into a grey and nebulous distance” (1964, p. 121). Notably, this distinction, given the language employed by Marx in the EPM, requires a charitable reading of his critique. However, the development of Marx’s alienation and critique of political economy extend beyond the EPM. Once we account for the evolution of Marx’s alienation in the full historical and philosophical context of his larger project, the objections outlined above lose their sharpness. This becomes especially clear when considering the sophistication of alienation as a material process throughout the Grundrisse.

The readings and reinterpretations of alienation described above remain valuable works in their own right; however, the wholesale rejection of alienation’s Marxian roots on the basis of its adherence to a perceived essentialism reflects a limited reading of the material from which it emerged. Recent work has applied Marxian or contra-Marxian alienations to examine various digital technologies which define contemporary life, including surveillance, social media, big data, artificial intelligence, and more (Haga 2022; Ashruf 2025; Sidorkin 2025; Bloor 2025; Küçükuncular 2025). These applications vary in methodology, approach, and form of alienation employed, with some situated purely in the scope of critical theory, and others within the application of quantitative and qualitative inquiry aimed at refining theoretical consideration.

2.3 Alienation and Machinery: Beyond the Economic and Philosophic Manuscripts

Recent reconsiderations have shifted alienation away from a problematic, perseverant punching bag of Marx’s young Hegelianism (Øverseen 2022; Zoubir 2018). Once a specter haunting critical theory, alienation has started to gain recognition as a foundational keystone of Marx’s thought. Zacharias Zoubir emphasizes the sheer importance of alienation’s development within the Grundrisse, though the text has largely been neglected in this right (Zoubir 2018, pp. 710–711). Throughout the Grundrisse, Marx reframes alienation as a process through which labor activities are appropriated away from the worker and concretized into fixed capital. Fixed capital refers to concentrated, unmoving stocks of invested capital that occupy a point in the process of production. Marx uses this term in reference to the large-scale industrial machinery which greatly transformed production around the time of his writing in London. Such machines come to appear as manifestations of capitalist production in a physically-embodied form. These machinic embodiments of capital greatly increase productive capacity, though they fail to liberate the worker from her subjugated position in production. This is because the incorporation of the machine into the capitalist organization is motivated by the increase of profit through the ever-growing speed, scope, and generalization of production. Capital remains the goal. As mechanization widens profit margins, it simultaneously widens the gap between worker and productive process.

Over the course of this development, utility and human engagement becomes supplanted by the surplus of capital generated under the increasing momentum of production. Under this momentum, labor exists in a constant flux of metamorphosis as it is continuously “reduced to a mere abstraction of activity…determined and regulated on all sides by the movement of the machinery” (Marx 1993, p. 693). Labor’s metamorphosis concludes, Marx contends, with the automaton, or automatic system of machinery (p. 692). Having outgrown its chrysalis, production emerges “merely as a conscious organ, scattered among the individual living workers at numerous points of the mechanical system” (p. 693). This transformation decidedly reshapes the epistemic and social milieu surrounding labor and production in general.

Production for purpose is replaced by mass production for profit. As Marx writes, mass production “destroys every connection of the product with the direct need of the producer, and hence with direct use value” (1993, p. 694). As machinery conducts an ever-increasing proportion of labor activity, it swallows up any associated technical know-how once required in the product’s manufacture. Having lost its initial context and meaning to the machine, the productive act, once conducted by human hands, is no longer perceived as originally human. Continuous machinic acquisition of production casts the activity of certain workers as inefficient, superfluous, or otherwise subordinate to that of the machine.

Imbued with the technique, power, and unparalleled speed of production, the machine comes to appear as the originating point of all expertise, knowledge, and value in production. Having alienated these capacities from the worker, the machine stands as an imposing body of production separate from, rather than complementary to the worker’s activity. Labor activity becomes deprecated to a system “of numerous mechanical and intellectual organs…the workers themselves are cast merely as its conscious linkages” (p. 692). The process of production no longer belongs to the workers, it has been appropriated away, and concretized into the form of the automaton. As the machine ascends on high for its originary expertise and uncontestable rate of production, the human is depreciated and deskilled. The machine, as the hulking, alien object, takes primacy over the newly-obsolete workers.

This form of machinery is notably distinct from the instrumentalized tool, which “the worker animates and makes into his organ with his skill and strength, and whose handling therefore depends on his virtuosity” (Marx 1993, pp. 692). Marx maintains however that machines are nonetheless the product of human ingenuity, functional “organs of the human brain, created by the human hand; the power of knowledge objectified” (Marx 1993, p 706). Taken on this basis, any and all technical production involving the instrumentalization of tools or the development of machinery is the result of an externalization of human capacities, and can constitute unalienated activity under the right conditions. This externalization itself occurs through the engagement of the brain and body to produce a tool directly, or in the use of that tool in pursuit of an end. The ultimate source of this production then, for Marx, is the human body, acting within nature, utilizing the materials at one’s disposal. However, such production in the abstract is incomplete without a greater investigation into that which underpins it.

3.1 Technicity and Self-Production: Grasping the Unalienated

Marx’s work in the Grundrisse is much less concerned with species-being, and much more concerned with the material organization and reorganization of production around the form “most adequate to capital” (Marx 1993, p. 694). However, Marx continues to gesture towards an unalienated standard which characterizes the direct mobilization of the body and tools in production, harkening back to his earlier concept of the species-being (pp. 694, 692, 706). Marx only invokes the term at one point throughout the entirety of the Grundrisse, and never presents an explicit conceptual continuity between species-being and the unalienated externalization of the body into technical instruments.

We may continue to reverse engineer some understanding of unalienated production through his reference to the tool-made-organ, coupled with a negation of his initial four-part diagnosis, though the resulting concept would hinge on a number assumptions and lack the explanatory power offered by the Grundrisse. It is tempting, then, to move on wholly from the EPM in favor of the Grundrisse; however, the earlier form of alienation outlined within the EPM nonetheless persists as the canon account within the larger body of Marxian scholarship. Further, the species-being stands to offer significant analytical strength if aptly defined, and remains crucial in illustrating the continuous developmental arc between Marx’s work in the EPM and the Grundrisse. What is missing in both pieces, however, is an elaborated analysis of the relations between humans and the technologies they produce through direct embodied activity.

This particular question of technological production and the human body — filtered through life in the developing hill country of Texas — would motivate fellow German exile, Ernst Kapp, to publish his work, Elements of a Philosophy of Technology in 1877. Kapp roots his account of technology within his concept of organ projection, situating the development of simple tools as outward externalizations of the body, echoing Marx’s nod towards instrumentalization. Kapp emphasizes that “as the human being makes use of the objects ‘at hand’ in its immediate vicinity, the first tools appear as extending, strengthening, and intensifying the human being’s bodily organs” (2018, p. 36). Kapp, similar to Marx, emphasizes the importance of the hand as the technical organ which served as the inborn prototypical model and source of all tools — from the primitive to the industrially and mechanically complex (Kapp 2018, p. 35).

Kapp centers the emergence of tools within a growing technical consciousness of the hand, which was refined through the development of primitive techniques conducted directly with the organ (p. 43). The use of primitive hands as tools, Kapp notes, likely produced a rougher, more utilitarian morphology (p. 35). However, through the creation of simple tools, these primitive organs were “relieved of sustained, immediate contact with raw, coarse matter” (p. 35). The development and continued use of increasingly sophisticated hand tools produced a more dextrous organ “by way of reciprocal action, the tool supported the evolution of the natural organ…in turn supporting, at each corresponding higher stage…the evolution and perfection of the tool” (p. 35). Kapp approaches primitive technological productions not solely meeting the material needs and creative impulses of early hominids. Rather, productions including weaving, fiber arts, the biface, the spear, and the hammer; intellectual productions of language, art, symbolic and phonetic writing, each represented a creative, dialectal expression of externalization and internalization. Kapp then describes this process within his explicitly dialectical framework.

3.2 Kapp’s Dialectic of Technology

Kapp differentiates between three stages within his model: (1) the rudimentary understanding of manufacture through direct use of the body; (2) the bidirectional impact of production upon material and body; and (3) the sophisticated understanding following observation and use of an artifact. It is in this third positive stage of the dialectic where knowledge of technique and mechanical law can emerge. Within this early stage of argumentation, Kapp employs a Hegelian, biological-material analysis, beginning with the primitive hand, and passing into primitive tools through an organologic relationship involving both body and environment.

Kapp fully elucidates his theory later as a “progressive, primarily unconscious process of the self-externalization of the subject, whose every act is not always at the same time becoming-conscious” (p. 53).8 Kapp characterizes this process as absolute self-production, or, an engagement in the productive act, through which the subject experiences a coming-to-consciousness, either alongside, or posterior to the act in itself. Kapp separates the process of externalization and that of coming-to-consciousness to account for the disjunction between knowledge necessary to fabrication, and any knowledge of processes implicit to the production or function of the tool itself. This knowledge may arise later in the positive stage of the dialectic. Here, the subject comes to understand the relation between the tool and the body, alongside the mechanical laws which govern all tools, primitive and complex.

Kapp points to the emergence of art and language as key examples of this process (p. 202). Kapp holds written language as “the greatest of all manufacts…an ideal, a work of art whose magic encompasses and records the whole human being...a spiritualized tool, both apex and agent of the human being’s absolute self-production” (p. 202). Advanced knowledge of linguistics was neither known, nor required for vocalizations to develop into language, nor language into symbolic inscription, nor inscription into alphabetization. Language emerged as a technique to externalize thought through significatory vocalization. Through this development, humans undertook a creative, exploratory effort, learning to use the mouth for communicative purposes, increasing problem-solving and coordinating capabilities for the purpose of survival. Early humans articulated internal sensations as emotions and expressed abstract thought in spoken, written, or symbolic means. As this dialectical process unfolded, language was instrumentalized as a tool and a means of interrogating our surroundings, experiences, bodies, tools, our own understanding, and language itself.

It is this very process of self-externalization through the organization of material, language, or thought that the subject comes to learn and understand humanity, environment, and the self. As Kapp writes, “from the tools and machines he has created, from the words he has thought up, the human being steps forth, the deus ex machina, face to face with himself!” (p. 250). Drawing on Kapp’s absolute self-production, we may begin to grasp what an embodied, unalienated production of tools and technologies looks like — notably abstract from the external, alienating organization of capital. Marx identifies a similar relationship in the Grundrisse, noting the process in which a worker uses tools as organs to leverage her virtuosity. Kapp’s account unveils the particularities of this relationship, advancing a vibrant characterization of embodied production in the process. Kapp applies his framework across the territory of the body, the organs, primitive tools, industrial machinery, and contemporary social formations — recontextualizing each as components which, taken cumulatively, represent their own technological organism.

In conceptualizing human technicity as an extension and externalization of the body, the following question arises: what differentiates ourselves from the machines we produce? Kapp, quoting Otto Liebmann, asserts that the organism emerges “ex ovo according to a hidden, immanent law” guided by an innate hegemonikon (“intelligence and will”) (p. 98).9 The machine, on the other hand, emerges ex homo. Consider a steam engine. When disconnected from a larger assemblage, it appears to lack certain capacities afforded to a human. Even while functioning within an assemblage, the engine exists as but one constituent part across a milieu of connections which enable its performance. Furthermore, the steam engine cannot set itself in motion. There remain gaps within its activity that require some form of human supplementation to fulfill, whether that is the incorporation of an additional machine or the activity of a human worker-as-conscious-linkage.

While a machine may carry out a certain encoded procedure, for Kapp, the essential hegemonikon of the machine remains with its fabricator and operator. Here we encounter another problem. What is to be said, then, of the organism’s steam-engine, the heart? What of the whole assemblage of organs that enable the activity of the organism? What of their manufacture and emergence ex ovo? Even in Kapp’s account, the delineation between machine and human relies on an appeal to some essential human quality. Despite this, two pertinent points of analysis may be drawn nonetheless: machinic function and originary emergence.

4.1 Human, Embryos, and Organ-Machines: Ruyer’s Model

Raymond Ruyer deepens Kapp’s notion of organ projection, approaching the brain, the body, and its organs as technological artifacts in themselves (Ruyer 2016, pp. 21, 55). Ruyer asserts that tools and organs are conceptually interchangeable as “both presuppose sense and finality, as much in their construction and constitution as in their use…the organs conform to the same norms of utility and of aesthetic and technical productivity” (p. 17). Ruyer stresses the interrelatedness and purposiveness of organs and externalized tools, as “it is through organs that man’s finalist activity can build tools and other works of culture” (p. 17). Finalist activity refers, then, to forms of free, senseful, engaged work-activity defined by an end (p. 11). Finality, as it relates to ends within work-activity, refers to the provision of sense towards establishing a virtual goal, in which one's work is creatively engaged in striving towards (pp. 8-11). It is through the judgement of the results of this engagement that value may be determined (p. 13). An act can be judged as successful or unsuccessful from any number of normative registers, which may then guide future activity (p. 13).

Organs and tools, while conceptually interchangeable and central to finalist activity, maintain distinguishing characteristics in their respective productions. Each organ has developed to a point of finality through an immanent, internal process of embryogenesis not immediately apparent to its possessor. A protozoa has no ability to observe and classify its organelles, nor does it hold knowledge over their function. Despite lacking a complete, systematic understanding over the conditions which constitute its production, the protozoa nonetheless instrumentalizes its organelles towards the ends of feeding and reproducing itself. While humans have come a long way in developing biological and scientific understanding more generally, any accumulated knowledge was advanced and collected through the direct production and use of tools. Ruyer importantly notes that the manner in which our tools are “built and inspected from the outside” contributes towards an epistemic predisposition for causal interpretation, guided by one's senseful investigation of the environment (p. 44).

Tools not only shape our bodies, organs, and brains, but also the understanding garnered through their use and proliferation. This is not to say that our organs cannot, or do not function within causal chains, nor that causal interpretation is wholly misguided. Rather, organs which have developed to a point of final instrumentalization are produced in a manner distinct from mechanical analogues built to reproduce the same processes carried out by organs. Organic, living tissue may retain self-regulative capacities, characteristics, or relations to other organs immanent to their larger behavior within the organism (p. 43). Thus the processes which unfold within the organism can be characterized, but not fully captured within a causal understanding to the same degree as a machine or technological analogue (pp. 43-44). While our investigative deployment of sense and technology has produced accurate mathematical and scientific knowledge of the real, there remain edge cases and blindspots where our instruments break down, and fail to produce the totalizing, cumulative understanding we often seek. This point reiterates the importance of extending interrogation beyond production alone, and towards the veiled relations which produce production.

Ruyer applies this heuristic across Neofinalism, extending his analysis even further when detailing the precise mechanisms of embryogenesis and technicity. Ruyer posits a sort of phenomenal primary consciousness lying at the root of the embryo. He asserts this primary consciousness precedes the sensory phenomena of the brain (2007, p. 6). For Ruyer, the brain is not the source of consciousness, rather, the brain developed through an exploratory process resulting in the objectification of consciousness into an organ which may process secondary phenomena.

Ruyer proposes that this primary consciousness remains active to some degree within all aspects of material, from molecular recombination to the genetic processes which ultimately shape embryonic development (2016, p. 110). The embryo, instrumentalizing the immanent production of genes, constructs sensory organs, which, united by the nervous system in the body, gives way to our secondary consciousness. This secondary consciousness emerges through a confluence of sensory phenomena derived from the body’s organs. Once the body is constructed, the brain takes on the role of the embryo, allowing us to interact with the environment; harnessing our sense towards engagement in finalist activity. Ruyer’s picture of human activity can thus be described as a secondary set of interactions which unfold across assemblies of organs and causally-motivated tools — constructed through that very activity enabled by that sense within the external world.

Ruyer’s work presents a movement away from the semi-materialist theorization of Kapp towards a neo-phenomenological material conceptualization. While not explicitly Hegelian, Ruyer’s emphasis on the primacy of consciousness echoes a Hegelesque interpretation of creation as being experiencing itself. Where Ruyer distinguishes his theory from Hegel, is in his rejection of absolution. Hegel emphasizes being as a divine process. Ruyer views being as a process of consciousness manifesting itself through its absolute material forms across chemical, biological, and neural processes. Importantly, Ruyer distinguishes the consciousness that exists within molecular interaction and organic life, asserting that the two are not linked by any underlying essentially divine cause. Rather, consciousness remains an emergent process, enabled through the proliferation of interactions between atomized components within the conditions of their external environment (p. 110).10

4.2 Marx, Kapp, and Ruyer: Comparisons, Considerations, and Limitations

Ruyer’s approach remains rooted within material, all the while centering the production of consciousness in place of the hand as the foundation of production and technicity. This consciousness manifests through its objectification in the construction of organs (like the hand) and the organism, the development of the brain, and through the organism’s brain and body back into the world. Organs achieve their finality in functioning as tools, giving way to the organism and, in the case of humans, allowing for the dialectic identified by Kapp to unfold. Ruyer complicates the third stage of Kapp’s dialectic, emphasizing that, while the mobilization of sense in tool-driven activity offers a coming-to-consciousness (namely of a judgement of value), the shape of that activity lends itself to the denotation of secondary, step-by-step causality. Sensory organs enable the external survey of processes and artifacts active in the body and the external world. Causal denotation is carried out through observation and in making “sense” of processes at play. The capacity to survey, assign cause, and to act with finality, is not essential; rather, it is an emergent, secondary capacity, enabled by the proliferation of sense.

Ruyer identifies and unravels the “hidden, immanent law” Kapp holds as the source of the organism, while Marx unravels the immanent forces which shape and transform production at the economic and societal levels (Kapp, 2018, p. 98). Both Ruyer and Marx emphasize the nondetermination of these processes. Ruyer importantly distinguishes the “senseful, free, valid, inventive activity” characteristic of finality with the “notion of a pure functioning, a pure succession of numberable causes and effects coming one after another…without any reversibility — a succession that is, by definition, incapable of surveying itself” (Ruyer 2007, p. 13). Marx similarly identifies a predilection towards uncritical causal survey of the “historical presentation of development” which, he argues, “is founded, as a rule, on the fact that the latest form regards the previous ones as steps leading up to itself, and, since it is only rarely…under quite specific conditions able to criticize itself…it always conceives them one-sidedly” (Marx 1993, p. 106). This view of historical development, in which all prior history has built cumulatively and progressively to present economic, productive, social, and technological formations reinforces its own necessity solely by virtue of its present existence.

A backwards-facing gaze towards all that precedes the present state of human development — whether directed at our organs, our productions, technologies, or the cumulative knowledge garnered through the lens of our tools — makes it tempting to argue the processes involved unfolded deterministically. It is this same gaze that lends itself to the essentialism of any form of production, especially towards the manner in which LLMs are produced and interpreted. It is this same gaze that, in a very real sense, imbues Kapp’s account of technological production with its essential hegemonikon. Taken on this basis, what is lacking most significantly in Kapp’s account compared to Marx is the critical consideration towards the predominant productive relations which come to preside over and shape the more general advancement of technology. Kapp asserts, on the basis of the hegemonikon, that technology will never advance beyond the capacities of the human, though he falls short of conceptualizing a reality in which technology confronts the human not as deus ex machina, but as an obstacle to absolute self-production.

Ernst Kapp advances a neutral, if not optimistic view of machine technology, while Marx’s view appears more pessimistic. This pessimism has been a source of derision among some technological inquiries into alienation, as Fabio Grigenti claims Marx was simply a “prisoner of an anthropology that saw limitations as values to be preserved and defended” (2018, p. 29). Grigenti dismisses the whole of Marx’s analysis on this basis, arguing that Marx was incapable of conceptualizing of, or conceding that some forms of technology could be leveraged towards a “path to self improvement” (p. 30). Contrasting Marx’s views with those of Kapp and Arnold Gehlen, Grigenti asserts a particularly strong reading of Gehlen’s deficient human (2018, pp. 33-36). Grigenti claims it to be “imperative for us to exonerate ourselves from a condition that is seen as impaired and inadequate;” positioning liberation solely on the plane of the individual body (p. 36).11

While Marx is absolutely a skeptic, his skepticism lies not with technology and machinery as a category in itself. Marx explicitly characterizes machinery as “products of human industry; natural material transformed into organs of the human will…or participation in nature. They are organs of the human brain, created by the human hand; the power of knowledge objectified” (1993, p 706). Marx also emphasizes the achievement and the liberatory potential offered by technology throughout the Grundrisse. Marx observed the transforming situation of labor in England and saw how machinery could be deployed to the distinct detriment of those that worked the factories. Marx thus held mechanization under capitalism — a system which channeled machinery towards the increase of generalized productive force at the expense of workers — as an unfortunate misapplication of this potential. Rather than liberating the worker from labor, necessarily enhancing her expression within it, or allowing her pursuit of self-production, she would continue to undergo subjugation, abstraction, alienation, and impoverishment.

5.1 Impacts on and of Production: Mechanization and Anti-Production

Interrogating capitalism with an awareness towards the system of logic it employs allows us to identify its causal boundaries, and observe an interruption of production on a scale at and beyond the human. Capitalism allows for certain forms of production, some of which can be constituted as creative, especially when judged from the normative standard of value it realigns production towards. There however remains a persistent difficulty in squaring “productivity” with the productive force advanced under a modern capitalist organization. Undoubtedly, the continued advancement of machinery and technology more generally have increased the capacity at which production is carried out. In this pursuit, however, entirely anti-productive strategies have become a defining characteristic of contemporary production.

One strategy of anti-production, planned obsolescence, sets out to engineer products with shortened lifespans — such that each is predestined to break down, become undesirable, and require wholesale replacement (Hamblet 2025). Replacement requires the purchase of another instantiation of the product, which, again, is predisposed to break down in a matter of time. Is this efficiency and innovation at work? When observed purely from the standpoint of capital generation, perhaps yes! Planned obsolescence is certainly an efficient and effective strategy towards minimizing manufacturing cost while ensuring the future inflow of capital. However, this case demonstrates a contradictory rupture between forms of production (or anti-production) which guarantee profitability and private good, and those which may hold utility and public good. Importantly, while there may not always be movement towards truly productive engagement on the record of capitalism, there remains a form of movement, that is, the movement of capital.

Whether the inward flow of capital created by a new market, the concretization of capital into new fixed forms, or the evaluative recording of gross domestic product — capital remains in, and judged by its constant motion. Drastic movements of capital upon this record stand to greatly impact the material and social realities it organizes. This can be seen with the degree of capital being poured into LLMs and other generative AI technologies today. The lack of extensive public knowledge of the function of this emergent, available technology — coupled with the promise that it is the innovative panacea of the century — generates a social momentum which feeds back upon the momentum of the capital flow driving its market forward. As a result, speculative investment and development accelerate further, cutting corners and encouraging more widespread adoption. This feedback takes differing forms across different fields.

Marx holds that, to the extent that a worker’s activity restricts or does not contribute to the continuous production of capital, or the production of production, that activity appears to be of little import. Similarly, if the activity involved in a labor act can be more or less replicated by an instantiation of fixed capital, then the mechanized alternative will be adopted — even if the commodity lowers in quality. Capital will continue flow, regardless.

5.2 Marx, Ruyer, and Deleuze: Causality, Capital, and Structural Alienation

In unleashing new flows of capital across production, capitalism has drastically increased access to more generally-consumable, desirable commodities. The industrial machinery of Marx’s time opened up worlds of possibility in production and trade. Likewise, the technologies that define the present have eased interpersonal communication, made information accessible at our fingertips, and drastically increased economic and social interconnectivity. However, each of these advancements — industrial, commodity, and digital — has been shaped and limited by the underlying epistemic and logical milieu of the system that organized its production. Marx outlines the implications of this tendency at the societal level more generally, noting that, across any productive epoch,

There is one kind of production which predominates over the rest, whose relations thus assign rank and influence to the others. It is a general illumination which bathes all the other colours and modifies their particularity…which determines the specific gravity of every being which has materialized within it. (1993, p. 107).

Deleuze and Guattari identify this tendency for one form of production to “take credit” for all production through the lens of what they call the socius and its respective surface of recording (2009, p. 10). Deleuze and Guattari assign the socius as a descriptor to capture the more generalized social organizational epochs which come to define production. They assert that production, consumption, and any judgements thereof occur upon the given socius’s recording surface.

Dan Smith notes one of the central carryovers from Marx’s work into Anti-Oedipus is the larger tendency of capital to move objective essences towards generalized subjective abstractions (Smith 2011, p. 39). As it relates to production, this tendency manifests through the development of any specified representation of labor into what they call “an abstract and deterritorialized subjective essence, the activity of production in general” (Deleuze and Guattari 2009, p. 270). Smith more straightforwardly identifies this general activity as “labour capacity…my capacity to produce” (2011, p. 39). It is through this generalization of labor capacity identified by Marx that production “become[s] the means of creating wealth in general…[ceasing] to be organically linked with particular individuals in any specific form” (Marx 1993, p. 104). Notably, the classical alienation, among a number of other concepts central to the traditional Marxist canon, are missing throughout Anti-Oedipus (Smith 2011, p. 38). However, alienation is not wholly absent from the text.

The alienation gestured towards by Deleuze and Guattari is approached largely as this process through which production reaches its generalized state, in which “abstract labor becomes something real” and, restating Marx, provides a standpoint “from which all the preceding social formations can be reinterpreted from the point of view of a generalized decoding” (2009, p. 302). In short, Deleuze and Guattari, accept, and directly advance the alienation of the Grundrisse through their centralization of production in general. They describe this trend towards generalization in terms of deterritorializations and reterritorializations, processes through which various social bodies undergo structural, logical, and relational transformations, ultimately becoming realigned to serve the ends of the presiding social organization (or socius) — which, in this case, is capitalism.

As the primary mode of global economy, capitalism is thus held to be the driving force of all production, past, present, and future. As a result, the logic of capital “falls back upon (il se rabat sur) all production” (Deleuze and Guattari 2009, p. 10). All judgements regarding what is productive arise directly from the production of capital. This logic shapes collective understanding of human development, as Marx noted within classical political economy. This logic similarly colors conceptualizations of present and future economic and technological trajectories, imbuing both with an implicit determinism.

This pattern is immediately apparent within the technological developments of the digital age. The internet, the smartphone, and machine learning, are but a few among countless other technologies that could not have been created without significant public investment into their development. Despite this, such technologies are almost solely accredited to the unyielding, innovative spirit and logic that lies within the invisible hand of the free market. As Deleuze and Guattari write, this logic claims:

To itself the whole and the parts of the process which now seem to emanate from it as a quasi cause…Forces and agents come to represent a miraculous form of its own power: they appear to be “miraculated” (miraculés) by it…Society constructs its own delirium by recording the process of production; but it is not a conscious delirium, or rather is a true consciousness of a false movement, a true perception of an apparent objective movement, a true perception of the movement that is produced on the recording surface. (2009, p. 10).

The miraculation of capitalist production, particularly within the field of technology, gives rise to the belief that all production trends firmly in a cumulative, forward direction. Herein lies the consummation between the more generalized, determinative, self-uncritical logic of causality, identified within Ruyer’s survey of tools and the body, and the imposed necessity of capital generation.

Marx, alongside Deleuze and Guattari, assert the result of this logic is that, over time, mechanized production comes to appear superior to the production engaged in by humans (Marx 1993, p. 694; Deleuze and Guattari 2009, p. 10). The same is true for technology in information and knowledge production. Data has historically held an ascendent sense of objectivity. The statistical figure is frequently perceived as distant, if not fully removed from human bias (Porter 1986, p. 5). While the reality of how data has been leveraged across history largely dispels this perception, it did little to harm its reputation, and did not slow its use as it developed into a tool of oppression and surveillance.

Similarly, LLMs have repeatedly demonstrated infiltration of bias, inaccuracy, and disinformation; yet their capacity to act as stochastic parrots has hardly impacted their deployment and adoption (Bender et al. 2021). Generative AI technologies differ from the industrial machinery of Marx’s time insofar as they remain relatively accessible and inexpensive to use. In this right, LLMs seemingly democratize the ability to produce specialized written content at a rate far beyond the individual at a fraction of the mental and labor cost. However, a critical survey of the manner in which generative technologies are produced and engaged with reveals the formative organizational gazes of capital and causality at play, giving way to the same tendencies of the industrial machinery identified by Marx.

5.3 Causality and Capital: The Production and Anti-Production in AI

When we examine the general shape of large language model interactions, a user enters a prompt, and receives a response. The prompt, as the cause, produces an effect in the form of a fluent answer derived from a thousand terabytes of data (Laizure 2024, p. 1).12 Upon reception of this cause and effect, we may be inclined to view the tool as a reflection of the same process which occurs within human interaction, writing, and knowledge production. We presume the tool fully understands the user’s prompt, and grasps the meaning of the response it produces. Inspecting this tool from the outside, such a presumption is not wholly misplaced. We have no other benchmark to compare our interactions with LLMs besides our own. Large language models, trained on petabytes of data scraped from the internet, are capable of providing summative answers for nearly any prompt a user may think up. It seems to follow, then, that more data will produce an objectively better model, one capable of grasping a more accurate understanding of reality and truth. If training data is continuously expanded, it seems we will inch closer and closer to the complete, totalizing font of knowledge we seek.

Large language models are capable of amassing and processing massive amounts of textual language data, though they notably lack the ability to comprehend or understand the conceptual meaning of the texts they produce. Emily M. Bender and Alexander Koller demonstrate this deficiency by delineating between two key components which make up human language. These are (1) linguistic form — the raw, written, digital, textual, or gestural data models are trained on — and (2) the situated meaning of language (Bender and Koller 2020, pp. 5185, 5187). Meaning is defined as the “relation between form and something external to language,” namely, communicative intent (pp. 5186, 5187). They argue that the training of LLMs purely on the prediction of words or strings does not simultaneously implant models with the capacity to learn, grasp, and comprehend the linguistic meaning, which persists between the word and its intentional context (2020, p. 5186). LLMs excel at identifying and establishing statistical relationships between tokenized words and phrases; however an LLM cannot derive meaning from form alone, no matter how large its sample is. Bender et al., paraphrasing Karen Spärck Jones, assert that blindly presuming the existence of understanding within LMs, or trusting that understanding will develop out of a continued expansion of training data imposes a restrictive set of methodological and epistemic assumptions (2021, p. 615). On this basis, researchers implicitly commit to invalid statistical relationships between inputs and outputs as reflecting accurate, objective relational meaning (2021 p. 615). Research thus abandons the pursuit of meaningful theoretical inquiry and understanding offered by language models, simultaneously flattening them into a source of convenience (2021 p. 615; Jones 2004).

The present, data-maximizing approach of large language model construction has been the subject of criticism for decades. Critiques largely stem from the resulting orientation of the field towards “state-of-the-art-results on leaderboards” which largely fail to build the “deeper understanding of the mechanism[s] by which they are achieved” as gestured towards by Jones (Bender et al. 2021, p. 610). Despite calls to slow down the expansionist approach and reorient the field to limited, rigorously-documented datasets necessary to guide technical knowledge towards developing Natural Language Understanding, datasets have only accelerated in growth (p. 610). A significant risk of hastily collected, multi-petabyte datasets is the accumulation of “documentation debt…a situation where the datasets are both undocumented and too large to document post hoc…Without documentation, one cannot try to understand training data characteristics in order to mitigate…attested issues or even unknown ones” (p. 615). As a result, the inner workings of language models have become more unwieldy, black-boxed, and opaque than ever before. This comes at a distinct disadvantage to the advancement of our own understanding in regards to this technology.

Language models are uniquely interesting technologies which hold significant potential when applied specifically and consciously. Research cohorts in the past have, and will continue to critique the present methods employed by AI industry giants, and justifiably so. They may propose more restrained approaches with an emphasis on data documentation, or interrogate the projection of understanding and consciousness onto language models. However, the ticket to organizing the production of the tool, its deployment, and any standards of operation lies with the capital those giants have at their disposal.

Language models have passed far beyond the point of documentation deficit. As a result, both the tools, and the market which drives the production of the tools are largely incapable of critical self-survey. Critical self-survey, however, is not necessary for the continued generation of capital. The present organization and production of large language models presupposes what Ruyer may call an instance of “pure succession of numberable causes and effects coming one after another” racing towards a necessary, predetermined, market-driven outcome (Ruyer 2007, p. 13). Even academic analysis in this field is not free from this optimistic, causal, techo-capital determinism.13

The present landscape of generative AI technology, by all accounts, appears dedicated to brute-forcing the appearance of understanding through an ever-expanding scope of blind, undocumented training data, with convenience as a guiding principle across development and deployment. Jones argued in 2004 that language models, regardless of appearances, remain bound by data volume and the fact that any production conducted by language models will reflect the average of that data (Jones 2004, p. 9). Jones argues that models will, at some point, reach a performance plateau which, in all accounts “may be quite high and may give the user quite a nice view of the surrounding landscape. But it will be somewhere on the side of the mountain, not on the top” (p. 9). Despite these limitations, we will continue to interpret meaning wherever its typical form is presented. LLMs process and present flows of information. While the interpretation and meaning flows from the individual reading the presentation, the appearance of conversational fluency within the technology remains convincing, as a recent survey found 67-percent of individuals attribute some degree of consciousness to Chat-GPT (Colombatto and Fleming 2024, p. 4).14 LLMs appear to many as agential, intelligent, independent third parties — presumably standing above human thought in objectivity and efficiency. The expansive, and ever-growing sets of training data further underline the perceived informational authority of these tools.15

In its constant opening of new markets and production of new commodities, capitalism seemingly provides workers with the potential to access new freedoms and opportunities. The most recent of these commodities being LLMs. The historical advancement of machinery and technology under capitalism has culminated in a transnational network of mass production with unprecedented interconnectivity and interdependence. Øverseen argues this development demonstrates the potential of capitalism to dramatically transform, increase, and potentially even democratize material and social power. Øverseen characterizes this stage of capitalism in terms of a dialectic of socialization and alienation in which one is confronted with “an objective reality that stands opposite the individual as an alien and unassailable force, and in which society itself appears as the greatest threat to human development and flourishing” (Øverseen 2022, p. 451). The attitudes surrounding the development of contemporary AI technologies exemplify this dialectic, and simultaneously showcase the process of realignment and deployment of social, scientific knowledge production towards forms most lucrative and adequate to capital production. Whether or not those forms are the most sophisticated, thoroughly understood, or thoughtfully constructed is generally superfluous if those forms can produce capital. The race to construct the one language model to rule them all has been driven, primarily, by the ability to generate the most amount of capital through tools trained on a maximal scope of data. This lies, in turn, on the presumption that enough data may give way to an optimal return on investment, albeit perfumed with the promises that these tools are a necessary source of salvation from global crises or the key to a liberatory future (See Pratt 2026; Sidorkin 2024).

5.4 Limitation and Productive Potential

While unalienated, absolute, or finalist forms of production remain possible under capitalism, the paths to achieving them are few, far between, and largely inaccessible. As a result, even the development of promising technologies become kowtowed by the necessity to accelerate productive force, negate necessary labor (now made obsolete), and return on investment. We must not forget the present driving force behind the productive arrangement of language models and technology lies with this manner in which capital incentivizes the organization of machinery and technology, not with the category of algorithms and machines in the abstract.

The present organization of this technology hinges on its ability to return on its speculative investments. As a result, its deployment is aimed at generating a situation which maximizes the production of capital through the generalization of mental production. A situation, in OpenAI’s own words “where intelligence is a utility like electricity or water and people buy it from us on a meter and use it for whatever they want to use it for" (Spirlet 2026). Through the generalization and “smoothing-over” of labor activity, that general activity of production, or labor capacity may be maximized in the widest possible cohort of individuals. This formulation of language models must be contrasted with the potential capacity of language models to function as instrumentalized tools or sources of inquiry, extending ability and even deepening understanding.

Kosmyna et al. note that, in the fourth session of their study, brain-only participants were provided access to Chat-GPT and LLM-group participants were restricted to the brain-only protocol. Across this session, brain-only participants “exhibited significant increase in brain connectivity across all EEG frequency bands when allowed to use an LLM on a familiar topic,” suggesting the “AI-supported re-engagement invoked high levels of cognitive integration, memory reactivation, and top-down control” (2025, p. 139). This result must be taken with the rather significant consideration that the participants in the brain-only group did not use LLMs prior to session four. It was through their pre-existing cognitive architecture that they managed to perform far above the LLM-to-brain group (p. 139). This further reflects the potential of technologies to act as instrumentalized tools, as well as their simultaneous capacity to act at the expense and alienation of the user. The positive and negative potential of our relations to technology is further demonstrated within Trond Haga’s 50-year micro-history of a shipyard corporation. Haga tracks variations in labor division and technology usage over the period, evaluating how shifts in work technologies transformed labor assignment procedures, work activity, and feelings of alienation among shipyard employees.

Haga identifies the determinant variables contributing to a sense of alienation as lying predominantly on the exterior (even outside the corporation). Senses of alienation were dictated by the degree to which technology increased or decreased worker accessibility to information and participation in decision-making (2022, pp. 812-813). More specifically, deprivation of “the opportunity to plan work and organize the execution” alongside a generalization of activities involved within labor were directly correlated with increased senses of alienation (p. 810). Embodiment within one's work was positively correlated with the degree to which technologies lended workers greater freedom and self-direction within their environment (pp. 810-812). Alienation soared when workers labored towards “individual operations and not a complete product” (p. 810). Haga’s findings, while derived from a very specific work scenario, reflect a similar sensibility of ownership over one’s effort identified by Kosmyna et al. When one is an active participant within one’s labor or activity, a sense of ownership over one's conduct is reflected within the attitudes towards that production. Haga’s study exemplifies the utility of quantitative, experimental approaches to bridging the theoretical alienation with its material manifestation as activities adapt to technological trends. The findings of this study further reflect the nuanced capacity for technology to either increase engagement and freedom within labor activity, or produce more alienating productive arrangements.

6.1 Towards a New Alienation: Creativity, Production, and Interruption

Capitalism seeks to generate value from production. In establishing itself as the prevailing normative standard of evaluation, capital is cemented as the guiding principle of all activity and production. Machines and technologies are constructed with the final purpose of capital in mind. As a result, their production, conduct, and relation to those who set them in motion reflect and reinforce the perceived necessity of this final purpose. From a Ruyerian perspective, under present conditions, we remain severely limited in our ability to generate new goals and virtualities to strive towards, insofar as those goals do not abide by the ascendant standard of judgement. While other value judgments may be established to guide activity in a very real sense, capital remains, at least in appearance, as the sole entry token to determine the shape of production on a large scale. In protecting existing markets, capitalism actively represses or wholly forecloses the dynamic and liberatory potential of its products through the anti-production necessitated by the continuous generation of capital. The result of this organization is a channeling-away of human labor, effort, and creativity towards profitable arrangements alone.

Creativity remains a significant aspect of the production humans engage in, though this capacity is not unique to the species-activity of the human. There exist creative solutions to unique and difficult problems present across nature — contained within genes and expressed through contingent interactions between organisms and environment. The application of creative thought towards the solution of a problem, or towards the ends of a purely creative endeavor, on the other hand, nevertheless remains a capacity that has historically shaped the continued existence of hominids. Ruyer’s articulation of finalist activity cuts across the grey and nebulous distance of systematic, totalizing causality which might frame this capacity as innate or essential. He achieves this through his identification of (often veiled) internal relations immanent to the production of the real, a tendency shared by Marx’s work. This argumentation is particularly useful to highlight across both Ruyer’s and Marx’s accounts, as it decenters an indelible view of the human as production’s unique starting point.

Marx argues in the EPM that the capitalist reorganization of production into the labor form alienated humans from regular engagement in creative productive species activity. The initial alienation of the EPM frames the foremost result of this shift as the replacement of direct subsistence through the environment and the community with reliance upon a wage for survival. It was this separation from the species-activity of self-production that yielded the state of alienation, worsened by the continuously degrading conditions in which labor was conducted. In this early stage of the concept we see Marx’s focus drawn primarily towards the causal effects of the relationship between workers and labor. Marx’s thought evolves as it turns to the mechanism through which the state of alienation was developed with his work Grundrisse, noting, and ultimately centralizing the process through which productive activity develops into ever more generalized, fixed forms. While the EPM presents an incisive and useful frame of analysis for understanding Marx’s early inquiries into capital proper, the Grundrisse represents an important developmental turn in his thought, and thus is vital in building an accurate understanding of alienation relevant to the present day.

Deterministic causality — whether implicit to the analyses of capitalism and human nature of classical political economists, the body and the production of technology by Kapp, or to the positive determinism of technology in the hands of market forces — stands opposite, and in the way of finalist activity, unalienated labor, and absolute self-production.

Alienation, in this formation, can be described as the process in which the labor, creativity, and effort involved in productive human activity becomes reoriented and channeled towards certain forms of production. Capital, as the normative standard of evaluation thus determines those forms. As a historical tendency, any form of production not already aligned to the production of capital becomes reoriented towards that end. The forms of production that stick around do so in light of their profitability. New movements that emerge under a capitalist organization may be creative in some manners, but above all, must contribute to the generation of profit. The necessity of profitability creates strictures to what forms of creativity and activity are judged as productive. While facilitating forms of production adequate to capital generation, capitalism suppresses other forms that may truly hold value when evaluated from another standard of judgement. Those judged to be productive warrant further investment, and those judged unproductive or extraneous either become concretized into machinery, or fall to the margins.

6.2 Towards a New Alienation: The Artifice of Machine Intelligence

A precise mechanism of alienation which accounts for generative AI technology is complicated to chart due to the presence of generative tools and content extending far beyond the scope of labor. An account of cognitive estrangement is further problematized by the seemingly voluntary nature of direct engagement with the tools. No one is necessarily obliged to use LLMs or other generative tools; however, the economic momentum behind their development has greatly accelerated their proliferation and compulsory use. The immediacy with which LLMs have been adopted across education and labor contributes towards the continued development of market and social pressures which cast these tools as viable, if not necessary shortcuts. A student falling behind on assignments under the pressure of a competitive academic ecosystem, an overworked teacher too exhausted to come up with a lesson plan, or someone experiencing mental health difficulties unable to afford therapy may all feel compelled to use an accessible application advertised to them as capable of alleviating their associated pressures.

No one is forcing these individuals to use generative tools in these instances, but they continue to confront the individual directly, or loom in the background as an option when the labor, task, or situation at hand appears too difficult to complete. The constant availability of AI tools coupled with the outpour of AI generated content has provided them a practically omnipresent position across the digital commons. Generative AI technologies remain a virtually inescapable presence across commonly used platforms — even those that initially served utilitarian purposes. While in almost all cases some capacity of choice remains, the weight of the task bearing down can be effectively alleviated through the use of generative tools, making them a tempting alternative.

This tendency to turn to shortcuts under pressure is further reflected in the MIT study; participants in the search engine and LLM groups were often “more inclined to focus on the output of the tools they were using because of the added pressure of limited time…therefore staying focused on copying and pasting content” (Kosmyna et al. 2025, p. 134). Delegation of cognitive labor to LLM tools, whether in school, work, or creative activity renders the person engaged in work as a passive facilitator rather than an active participant. As a passive facilitator, one “merely transmits the machine’s work. the machine’s action” and “supervises it and guards against interruptions” (Marx 1993, pp. 692). In transmitting prefabricated LLM-generated outputs, individuals are less inclined to search for, digest, and interpret information themselves, often taking the output at face value (Kosmyna et al. 2025, p. 134). Passive facilitators are also less likely to synthesize and incorporate personal perspectives and insight within the content being transmitted (2025, p. 134). This lack of personalization and engagement is further substantiated when considering the homogeneity of content across LLM-group responses (2025, p. 133).

Despite the loss of direct engagement, one advantage of LLM use that cannot be underwritten is the associated increase in efficiency compared to traditional manners of engagement. A 2021 study reported participants demonstrated a 60% increase in overall productivity with LLM-guided application use compared to direct use (Schmidhuber et al. 2021, p. 4). This boost in productivity was similarly tied to a lowered cognitive load across varied tasks (p. 4). LLMs present an unprecedented means through which users may access information, reach a deadline, and mitigate any stress that comes along with it. LLMs also lower frustration and mental load required by manual filtration and synthesis of information. This decrease in mental load notably increases the length users are able to work on a given task.

By all accounts, when compared to the strategies employed by the search engine or brain-only group, the mechanized alternative appears to be the more expedient option. Gone are the days of all-nighters, late submissions, and cognitive grunt work. Such a prospect undoubtedly appears as an immense, liberatory advantage. Workers, educators, students, business owners — anyone — can spend less time engaged in the excessive demands of work, research, comprehension, and writing. In these cases, however, the boost in efficiency proceeds directly from offloading engagement with the task to the language model. In lessening cognitive load, participants shift from “active critical reasoning to passive oversight” (Kosmyna et al. 2025, p. 15). The crux of such a shift lay in that, compared to the production of the other two groups, the writing produced by the LLM-group was neither owned, meaningfully known, or embodied by its participants. For all intents and purposes, when engaged in LLM-assisted production, the written word — the result of cognitive labor — is no longer the direct product of the individual tasked with producing it.

Observed from the standpoint of increased efficiency in productive capacity, LLMs seem to win out completely. This is because LLMs, as Marx might have it, represent the realization of a core tendency of capital to “increase of the productive force of labour and the greatest possible negation of necessary labour” (1993, p. 693). The means most adequate to increasing productive force — and in turn the production of capital — in this case, is the mechanization of knowledge production. In separating the production of written content or knowledge from the individual producing it, the necessary labor, skill, and education required for research and writing is seemingly made obsolete. It is precisely due to the ever increasing speed and scope of this general productive force that AI tools are reified as necessary implements to “keep up” with the times, or risk falling behind. LLMs do not necessarily free the individual from a greater number of tasks, rather they allow for a minimization of “the amount of labour necessary for the production of a given object…but only in order to realize a maximum of labour in the maximum number of such objects” (1993, p. 701). The process of LLM-driven production and its widespread deployment across institutions does promote a more efficient form of production — one capable of keeping up with increasingly tight deadlines across labor and education. However, the efficiency of this new arrangement stems from a diminished instrumentalization of the brain. Underutilization of the brain breeds reliance upon the tool supposedly made to supplement its labor.

While these tools appear truly valuable in “accelerating the process” as it were, and remain capable of producing relatively well-performing material, the results of the MIT study show any benefit existing squarely within the short term. LLM group participants displayed clear, negative, long-term hindrances in cognitive ability. This disadvantage, researchers note, is not merely behavioral, it “is mirrored in the structure and strength of their neural connectivity,” resulting in overreliance as well as “less sophisticated reasoning and weaker argumentation” (pp. 140, 17). While the use of a generative AI tool may relieve some degree of labor or stress in a pinch, deferring mental labor in the long term, even in small amounts, vitiates the brain as cognitive debt compounds. This result demonstrates an edge case of Kapp’s dialectic, in which the bidirectional relationship between organ and tool yields a less refined organ.

In removing the brain from “sustained, immediate contact with raw, coarse matter,” the cognitive labor is not enhanced, extended, or improved; rather, the brain is largely disengaged and diminished (Kapp 2018, p. 35). While cognitive debtors may exhibit short-term benefits in productivity, offloading labor to LLMs over extended periods “diminished critical inquiry, increased vulnerability to manipulation, [and] decreased creativity” (Kosmyna et al. 2025, p. 151). As Marx aptly puts it: “the more the work manifests intelligence the more the worker declines in intelligence…it produces intelligence, but also stupidity” (1963, p. 124). With all this in mind, the race to incorporate generative AI tools to as great an extent as possible across education, labor, and everyday life — given its hindrance to critical thought, creativity, and independence from LLM use — appears to be a race to the bottom. LLMs and generative technologies cannot be understood as a necessarily forward-facing, purely beneficial technological advancement in the long successive, causal construal of the brain’s evolution. Rather, the manner in which these technologies are produced presently reflect the general progression of scientific production, albeit interrupted and preemptively deployed in a commodified form adequate to capital, aimed at increasing the general speed of production.

In the general progression of capital, whether throughout the industrial revolution or in the present, productive knowledge is continuously appropriated and concretized into fixed capital machinery. As a result, labor conducted by workers may no longer require the degree of expertise, effort, or time once necessitated by whatever product is the object of their labor. From a distance, this may seem a liberatory process in which opportunity is opened and freedom is seemingly produced for a wider number of potential workers. However, as Marx asserts, while the necessary effort and expertise required of labor is reduced to a minimum, it is done with the express purpose to maximize input and output of production. In transferring certain labor activities to machinery, workers lose out degrees of freedom, creative self-direction, and determination over the shape of their labor.

Despite consistent abstractions and alienations from labor activity, workers have largely remained species-beings capable of conscious and creative thought. We presently occupy a point in which there exist machines capable of mimicking the generation of creative and conscious thought, albeit through non-conscious means (Li 2025).16 Thought remains an untapped market of production, in which data may be gathered on the digital blueprint of written thought online, and any production thereof may be passed on to machines. While LLMs and other generative technologies allow their users to complete tasks much more quickly than it might take to develop the skills previously required by mental production, the fruits of this production become flattened and univocal (See Kosmyna et al. 2025, p. 133). In opening the door to large-scale cognitive off-loading, the AI models that come to mimic the abandoned processes of thought appear to think much more efficiently and effectively along the lines of their encoded logic — far better than the cognitively-off-loaded individual. This is not a sign of superiority of machinery over individual capacity, rather, the result of the internalization of the productive relations represented by technologies which facilitate the off-loading of productive activity.

There is no question of whether viable production takes place in the use of generative AI tools. A product is produced. The manner in which they are deployed, however, actively unseats vital cognitive capacities, and thus usurps the presence of self-direction within production further. The results of this usurpation do not appear to reinforce the utopic, liberatory potential often ascribed to them. Rather, generative AI, in its present formation, appears as the newest imposition of the logic of capital, confronting, and falling back upon the individual as necessity.

7 Conclusion

The approach to alienation outlined in this paper highlights the process by which the capitalist mode of production harnesses, and deprecates self-production into the alienated form as described by Marx. The absolute production of the self — characterized by Kapp’s dialectic of tools and organ projection — consists of the externalization of the body’s organs into technical artifacts. Through the dialectical model, Kapp differentiates between three stages: the rudimentary understanding of manufacture through the direct use of the body, the production of the artifact through the use of material, and the more sophisticated understanding that emerges following observation and use. The capitalist mode of production, as Marx notes, functionally deconstructs the whole epistemic thrust of Kapp’s productive model, restricting and concentrating knowledge into machinery, while replacing the fruits of the productive act with a wage. While this account is incredibly complementary to Marx, Kapp remains incomplete on his own, and fails to provide a rigorous defense against essentialism.

Ruyer’s work deepens Kapp’s dialectic by extending analysis down to organs, cells, and molecules. Creativity and technical production, in this view, need not be considered essences unique to human activity. Rather, conscious production and creativity permeate the abiotic and biotic worlds, with the former giving rise to the latter. In understanding the human organism through this lens, we can better understand how the modern and contemporary organizations of labor may interrupt or hinder truly creative productive engagements. While there undoubtedly exist structures and systems not inherent to capitalism that may restrict or interrupt engagement in unalienated activity, absolute self-production, or finalist activity, capitalism — in restricting movement upon its recording surface — greatly prunes the avenues and opportunities for workers to engage in the creative self production identified by Marx, Kapp, and Ruyer. Generative AI tools, by virtue of their present stage of development, production, and adoption, stand as a new stage in further entrenching the interruption of productive activity; casting human engagement in writing and the production of thought as obsolete, unprofitable, and flawed. While the full ramifications of cognitive estrangement brought about by this most recent mechanism of alienation are yet to be seen, the encroachment of these tools into nearly all institutions and facets of contemporary life demands proper consideration in this right.

Our activities, construction of tools, and making of meaning are often driven by external observation, societal norms, causal understanding, and their combined application towards subsequent developments and activities. Such external observation is useful, but must be interrogated with priority given towards unveiling conditions of production which may not be immediately apparent. This sentiment motivates Marx’s work as it relates to capital, Kapp’s work as it relates to technology, Ruyer’s approach to the body, consciousness, and tools, as well as Deleuze and Guattari’s diagnosis of capital’s logic. Any attempt to square the impacts of the generative technologies of the present must take care to approach the topic with this same critical rigor. Instead of falling into the promises of technological and economic determinism, inquiry must remain rooted in the present material reality, and direct itself towards critical analysis of the economic, social, ideological, and productive forces at work.