No Outside Permanent Exploitation in the Age of Surveilance Capital
1 Introduction: Permanent Exploitation
The nineteenth-century census counted you. The twenty-first-century algorithm becomes you. This is not a rhetorical flourish but the central analytical claim of what follows. The distinction between descriptive statistics and extractive data is not a matter of degree but of kind—a qualitative rupture that Marxist theory is uniquely positioned to diagnose, and one that most liberal critiques of “surveillance” systematically occlude.
There is a seductive narrative that presents the smartphone, the targeted advertisement, and the data broker as the latest iteration of an ancient bureaucratic impulse: the sovereign counting his subjects, the demographer mapping his populace, the insurer pricing his risks. Under this account, we are simply witnessing more efficient information processing. The story is false. What began as the state’s effort to render its population legible has mutated, under the pressures of late capitalism, into something Marx himself did not live to see but whose skeletal structure he described with prophetic clarity: the total subsumption of life under capital, and the emergence of a new category of surplus he had no occasion to name.
The argument advanced here is that capitalist exploitation has become permanent. The nineteenth-century worker could, at least in principle, leave the factory at shift’s end and cross a threshold into a sphere that capital did not directly valorize. That threshold no longer exists. The intimate and private sphere—the bedroom, the body, the nervous system, the half-formed desire at 2:00 AM on a Tuesday—has been converted into a continuous site of extraction. What is extracted there is not labor in Marx’s classical sense but something Marx could not have theorized, because the technical conditions for it did not exist: behavioral surplus, the exhaust of ordinary experience rendered into predictive commodities traded on behavioral futures markets as if they were financial derivatives.
To understand the scale of this mutation, we must trace a theoretical lineage that runs from Marx’s industrial baseline, through the Italian Operaist diagnosis of the “social factory,” through Gilles Deleuze’s analysis of the control society and the dissolution of the individual into the “dividual,” to Shoshana Zuboff’s mapping of surveillance capitalism. Each stage represents a deeper enclosure: of the working day, of social life, and finally of the interior itself. Each represents a further refusal of the boundary between capital and life. The essay closes on what I take to be the decisive theoretical claim: data is no longer information about production; it has become a moment of production, and indeed a form of fixed capital. This is what it means to say that exploitation has become permanent.
2 Part I: Marx, the Working Day, and the Hidden Abode
In the opening chapters of Capital, Volume I, Marx performs a pivotal methodological maneuver. Having walked the reader through the “noisy sphere” of circulation—the market, with its apparent freedom and equality—he abruptly pivots. “Let us therefore,” he writes, “in company with the owner of money and the owner of labour-power, leave this noisy sphere, where everything takes place on the surface and in view of all men, and follow them both into the hidden abode of production, on whose threshold there stares us in the face ‘No admittance except on business.’” The factory, not the marketplace, is where the secret of surplus value is manufactured. This methodological gesture is foundational: the real relations of production are hidden from the sphere of circulation, and the analyst’s task is to penetrate that concealment.
For Marx, the nineteenth-century working day is the central site of struggle. Capital, having purchased labor-power for its daily value, seeks to extract from it the maximum quantity of labor. Marx describes capital’s drive in the language of horror: it “vampire-like, only lives by sucking living labour, and lives the more, the more labour it sucks.” Elsewhere in the same chapter he names this drive a “were-wolf hunger” for surplus labor—a compulsion that consumes the time the worker needs for meals, rest, fresh air, family life, and what Marx calls “the time for human education, for intellectual development, for the fulfilling of social functions and for social intercourse, for the free-play of his bodily and mental forces.” This is the extraction of absolute surplus value: the lengthening of the working day itself, the consumption of human time and human bodies within the spatial container of the factory.
But note three features of this account that will matter decisively for what follows.
First, the factory has walls. The working day has a beginning and an end. When the exhausted worker staggers home, crosses the threshold of the domestic sphere, and collapses into whatever private life remains available to him, capital’s direct extractive hand is—in some meaningful sense—stayed. The home in the nineteenth-century industrial imaginary is a compromised and impoverished space, yes, but it is not the factory. The factory gates and the shift clock together constitute a real, if porous, boundary. The worker escapes. Not fully, not forever, and not without returning tomorrow—but the escape is structurally available. There is an outside.
Second, data in this regime is descriptive, not extractive. The nineteenth-century census measures the population in order to govern it. The factory inspector’s report—those Blue Books that Marx mines so mercilessly in Chapter Ten—documents the length of the working day in order, eventually, to regulate it. The state’s statistical apparatus is concerned with making populations legible, with counting hands, classifying occupations, measuring the misery of the industrial slums. This is the cartography of labor, not its real-time harvest. The census, as an instrument, measures the map. It does not clear-cut the territory.
Third, and most critically for the Deleuzian argument to come, Marx’s analytical subject is a class aggregate. The categories that matter to nineteenth-century political economy—industrial weaver, landowner, factory hand, rentier, bourgeois—are categorical, not individual. The worker’s specific psychological quirks, his particular rhythms of desire and inattention, the peculiar timbre of his voice or the dwell-time of his gaze on a shop window, were invisible to capital because they were irrelevant to capital. Labor-power was purchased in standardized units of time, and the worker was fungible within his category. What capital required was a class of laborers, not this laborer; a mass of consumers, not this consumer. The individual’s interior was noise, not signal. The nineteenth-century census recorded occupation, household size, literacy, and place of residence. It did not record, because it could not record, what the subject was feeling at 2:14 on a Tuesday afternoon.
These three features—the existence of an outside, the descriptive function of data, and the categorical nature of the analytical subject—define the classical regime of industrial capitalism. Each will be dismantled in turn by the four mutations the rest of the essay traces: the social factory’s dissolution of the outside, the dividual’s dissolution of the categorical subject, the cybernetic loop’s dissolution of the map-territory distinction, and behavioral surplus’s dissolution of the boundary between the interior and the commodity.
3 Part II: The Social Factory and the Dissolution of the Outside
The first mutation is the dissolution of the outside. This is the contribution of Italian Operaismo, and specifically of Mario Tronti, who in his 1962 essay “Factory and Society” advanced a claim whose analytical consequences we are still working out: that capital had broken out of the factory and colonized society as such. “At the highest level of capitalist development,” Tronti wrote, “the social relation is transformed into a moment of the relation of production, the whole of society is turned into an articulation of production; that is, the whole of society lives as a function of the factory and the factory extends its exclusive domination to the whole of society.” The factory had metastasized. Society itself had become the social factory.
Tronti’s insight was technical as much as political. Once capital had begun to extract relative surplus value at scale—once productivity gains, rather than the simple lengthening of the working day, became the primary engine of accumulation—the entire social infrastructure that produced the worker began to matter to capital in a direct, operational sense. Schools, hospitals, transit systems, mass media, cultural industries: all of these were now moments in the production of labor-power. The “outside” of the factory was, properly understood, simply another part of the factory, organized by different mechanisms but serving the same accumulative logic.
Antonio Negri and the post-Operaist generation that followed extended this argument in a direction Tronti himself had only gestured toward. In Empire (2000), Negri and Michael Hardt argued that contemporary capitalism had developed a form of “biopolitical production” in which the raw material of valorization was no longer merely muscle-power but the totality of social life—relationships, affects, knowledge, communication. Maurizio Lazzarato, in his 1996 essay “Immaterial Labor,” sharpened the point further: capitalism had come to rely on forms of work that produced “the informational and cultural content of the commodity,” including the work of shaping tastes, defining cultural standards, and producing the very subjectivities that would later consume the resulting products. “Immaterial labor,” Lazzarato argued, “produces first and foremost a ‘social relation’"—and, crucially, its raw material was “subjectivity itself.”
What this line of analysis establishes, and what matters for our purposes, is that by the late twentieth century the boundary Marx had drawn around the factory had become analytically indefensible. Capital had breached the factory walls. The commute was work. The television was work. The pub, the gym, the classroom, the bedroom—each was now a moment in the circuit of valorization, because each contributed to the production or reproduction of the labor-power that capital required, or to the formation of the subject capital required to desire its commodities. The outside that Marx had preserved, however compromised, was gone. The worker did not go home at the end of the shift; he went to another part of the factory.
But notice what this argument, in its Operaist form, still preserves. Tronti, Negri, and Lazzarato are theorists of the social, of the collective, of the mass subject. Their analytical unit remains, in the final analysis, a class subject—expanded, certainly, beyond the industrial proletariat to encompass the “social worker” or the “multitude,” but still a categorical subject. The factory had become society, but society was still analyzable in terms of class aggregates. The question of what happens when capital ceases to address the subject as a member of a class, and begins instead to address each subject as a unique, granular, ever-updating bundle of data points—this is the question the Operaists did not fully pose, because the technical infrastructure that would make such an address profitable did not yet exist.
For that, we must turn to a philosopher writing nearly three decades after Tronti and a few years before the World Wide Web.
4 Part III: Deleuze, the Dividual, and the Control Society
In 1990, in a short but decisive text titled “Postscript on the Societies of Control,” Gilles Deleuze proposed that the disciplinary society Michel Foucault had anatomized—the society organized around the factory, the prison, the school, the hospital, each a distinct enclosure with its own interior logic—was giving way to something new. The disciplinary society produced individuals by passing them through successive molds: the family molded you, then the school, then the factory, then (if you were unlucky) the prison. Each institution remained analytically separate. Each produced a recognizable subject: the pupil, the worker, the patient, the convict.
The control society, by contrast, produces not individuals but dividuals: no longer discrete subjects to be molded, but continuously modulated data streams to be read. “We no longer find ourselves dealing with the mass/individual pair,” Deleuze writes. “Individuals have become ‘dividuals,’ and masses, samples, data, markets, or ‘banks.’” The factory, which had once been a bounded site that produced identifiable workers, gives way to the corporation, which is “a spirit, a gas”—a modulating force operating continuously across time and space, pricing and repricing, scoring and rescoring, never settling on any final mold.
The distinction is not rhetorical. It is the precise theoretical hinge on which the transition from industrial to surveillance capitalism turns. The nineteenth-century worker was legible to capital as a member of a class: an industrial weaver, a coal hewer, a domestic servant. His individual quirks were noise that the accounting system discarded. His labor-power was standardized, and his subjectivity was, from capital’s perspective, essentially invisible. Even the Operaist “social worker” of the late twentieth century retained this categorical legibility; he or she remained a member of a recognizable class subject, however expanded.
The dividual is something else. The dividual is not a member of a class but a unique, continuously updated signature: heart rate at 3:47 AM, dwell time on the third photograph in a scrolling sequence, GPS coordinates correlated with purchase history, micro-expressions registered by the front-facing camera, voice-tone analysis during a customer-service call, typing cadence, sleep patterns, menstrual cycles, route deviations, search-query revisions, the specific seven seconds of hesitation before you click “buy.” The marketing apparatus does not care, in any meaningful operational sense, about your class position. It cares about your specific psychological vulnerabilities at 2:00 PM on a Tuesday. It cares about what will make you, this dividual signature, click now.
This is the second mutation, and it is as radical as the first. The nineteenth-century extraction regime was categorical; the contemporary regime is granular. The nineteenth-century subject was visible to capital as a type; the contemporary subject is visible as an ever-shifting cloud of data points. What had been noise—your individual rhythms, preferences, vulnerabilities, the idiosyncratic texture of your interior life—has become the primary signal. Capital has built an apparatus of such analytic resolution that the individual subject, precisely in his or her uniqueness, has become the operational unit of extraction.
And note what this does to the classical Marxist categories. Marx’s analysis of exploitation was calibrated for a regime in which workers could be organized as a class because capital addressed them as a class. The shared condition was visible; the solidarity was structurally available. Under the regime of the dividual, the shared condition is systematically disaggregated. Each user is presented with a personalized feed, a personalized price, a personalized advertisement, a personalized stream of political outrage. The granularity of the extraction is simultaneously the atomization of the subject. The dividual cannot easily recognize fellow dividuals as sharing her condition, because her condition is, by design, not shared in any immediately visible way. This is not a bug; it is the operational principle of the control society.
5 Part IV: Zuboff, Behavioral Surplus, and the Cybernetic Loop
With the ground thus prepared, we can finally make sense of Shoshana Zuboff’s diagnosis of surveillance capitalism, and of its properly Marxist significance. Zuboff’s central concept is behavioral surplus: the exhaust data produced by our ordinary interactions with digital systems—what we click, what we linger on, how long we read, where we go, what we search, whom we message, how our voices sound, how our faces look—which is captured, refined, and processed into predictions about future behavior. These predictions are the commodity. They are sold on what Zuboff calls “behavioral futures markets” to advertisers, insurers, political campaigns, employers, and anyone else with an interest in shaping or anticipating human action.
Behavioral surplus is not a rebranding of an older category. It is something genuinely new, and its novelty deserves theoretical emphasis. Marx, writing in the 1860s, identified two forms of surplus: absolute surplus value, extracted by extending the working day, and relative surplus value, extracted by intensifying the productivity of labor within a given day. Both are extracted from labor-power inside the wage relation. Both presuppose a worker who has sold a definite quantity of time to a capitalist, and who is producing value within that definite quantity. Behavioral surplus breaks this structure entirely. It is extracted from activity that is not sold, not waged, not understood by the subject as labor, and that takes place outside any recognizable employment relation. The scrolling teenager, the commuter checking email, the sleeper whose smart-watch records his pulse: none of these have entered into a labor contract, and yet all of them are producing behavioral surplus from which capital derives predictive commodities. Marx’s categories do not quite fit, because the technical conditions Marx analyzed no longer obtain.
This is why I insist on treating behavioral surplus as a third form of surplus, supplementary to absolute and relative surplus value rather than reducible to them. The digital regime has produced an extractive mechanism for which Marx’s conceptual apparatus was not designed, because the mechanism itself did not exist in his day. Faithful Marxism here requires not pretending that behavioral surplus is “really” just another form of relative surplus value, but recognizing that capitalism has mutated in a way Marx could have anticipated only in the most schematic terms. The werewolf has grown a new appetite.
The third mutation follows directly, and it concerns the temporality and function of the data itself. Nineteenth-century data was a map. It was descriptive, periodic, and slow. A census was conducted every ten years; a report was published; a royal commission deliberated; a bill was eventually drafted; in due course, perhaps a law was passed. The feedback loop between measurement and intervention ran in decades. Data was produced after the fact and was used to describe a territory that had already been constituted by other forces. It did not, and could not, constitute the territory itself.
Contemporary data operates on an entirely different temporality and serves an entirely different function. It is produced continuously, processed in real time, and fed back into the system in order to reshape the behavior it is measuring while that behavior is occurring. This is a cybernetic loop in the technical sense: a feedback circuit in which output continuously modifies input. When the platform detects that you have slowed your scroll on a particular image, it serves you more images like it, not in order to describe your preferences but in order to deepen them—to herd you toward patterns of engagement that maximize the platform’s extraction of behavioral surplus. The algorithm does not map your desires. It manufactures them, in real time, and then sells predictions about the desires it has itself instilled.
The difference between these two regimes of data is the difference between a map and an engine. A map describes a territory that exists independently of the map. An engine reshapes the territory as you walk through it. The nineteenth-century census was a map. The contemporary algorithmic apparatus is an engine. To collapse the two under the common label “data collection” is to miss the entire theoretical point. What surveillance capitalism has done is not to collect more information than its predecessors; it is to construct a machine that uses real-time information processing to continuously reshape the very behaviors it is measuring, in the service of continuously extracting behavioral surplus from those reshaped behaviors. This is a qualitatively different operation from census-taking, and calling the two by the same name obscures rather than illuminates.
The fourth and culminating mutation concerns the status of data itself. In the nineteenth century, data was a use-value: a tool for the state, a resource for the theorist, an instrument of administration. It was not capital in the Marxist sense, because it did not enter directly into a valorization circuit; it remained exterior to M—C—M′. The census produced information about the production process, but it was not itself a moment of the production process.
Under surveillance capitalism, data has become capital in the strict Marxist sense. It is now a form of means of production—arguably a form of fixed capital—that participates directly in the extraction of surplus value. Behavioral data is accumulated, refined, and processed into prediction products that are then sold on behavioral futures markets. The accumulation of behavioral data is the condition for the further accumulation of behavioral data, because larger data sets produce better predictions, which command higher prices, which fund the infrastructure that captures still more data. This is a self-expanding valorization circuit—capital in motion—and its raw material is human experience itself.
Marx’s census data never sought to claim ownership over the inner desires of the worker. It measured his caloric intake, his hours, his literacy, his household. It left his interior life alone, not out of respect but out of technical impossibility and economic irrelevance. Contemporary data collection claims precisely what Marx’s census left untouched: the interior. Your private experience—your desires, your hesitations, your late-night searches, your intimate communications, your physiological responses to the stimuli engineered to produce those responses—has been rendered as commodity and traded as if it were a financial derivative. The private sphere has not merely been surveilled. It has been financialized.
6 Conclusion: No Outside
Return, then, to the distinction from which we began. The nineteenth-century census measured the map. It was descriptive, periodic, bounded, and categorical. Its purpose was to render the population legible to the state for purposes of governance, taxation, and conscription. It addressed its subjects as members of classes. It left the individual interior untouched, because the individual interior was both technically inaccessible and economically irrelevant. Most importantly, it presupposed the existence of an outside: a sphere of private life that, whatever its miseries, was not itself a direct moment in capital’s valorization circuit.
Each of the four mutations we have traced has dismantled one of these features. The social factory dissolved the outside. The dividual dissolved the categorical subject. The cybernetic loop dissolved the map-territory distinction. Behavioral surplus dissolved the boundary between the interior and the commodity. What remains is a regime in which every moment of waking (and indeed sleeping) experience is a potential site of surplus extraction, every idiosyncratic detail of the subject’s interior is potential raw material, every act of apparent self-expression is potential input to a behavioral prediction that will be sold to the highest bidder. This is what it means to say that capitalist exploitation has become permanent: not that it is more intense than it was in the nineteenth century—in many respects, of course, it is less physically brutal—but that it has eliminated the threshold beyond which one could, however briefly, stand outside it.
The classical Marxist category of absolute surplus value described capital’s hunger to extend the working day toward the physical limit of twenty-four hours. In the nineteenth century, this hunger was checked by physiological necessity, by legal regulation, and by the existence of the home as a space of non-valorization. Under surveillance capitalism, none of these checks operate as they once did. The twenty-four-hour working day is not a rhetorical projection; it is the operational reality of any smartphone-mediated life. You do not clock out. The feed does not pause. The smart-watch does not sleep when you sleep; it records your sleep, and the record is sold. The werewolf has no gates to return to, because the werewolf now lives inside the house, and indeed inside the skull.
The strictly Marxist task, then, is to name this process accurately. To call it “privacy violation” is to misdescribe it as a rights issue when it is a production relation. To call it “data collection” is to naturalize it as benign bureaucracy when it is the construction of a wholly new extractive apparatus. It is, instead, the most audacious enclosure in the history of capital—a fencing of the interior itself, the final conversion of human experience into a resource for valorization, and the terminus (for now) of a long movement by which capital has progressively dissolved every threshold that once separated it from life. The hidden abode Marx entered in 1867 has been progressively dragged into the light: first the factory, then the entirety of the social, and now the psyche. Behavioral surplus is the name of the value form that corresponds to this final enclosure. It is a form Marx could not have theorized, because the technical conditions for it did not exist in his day; but it is recognizably continuous with his framework, an extension of his categories rather than a departure from them, and it demands a Marxist analysis that is willing to extend the master’s categories rather than merely to repeat them.
Recognizing this is the first step toward contesting it. The question for our century is whether the dividual can be reassembled into any recognizable collective subject capable of refusing an extraction regime that is, by design, calibrated to be invisible to its targets. That is a political question, and nothing in Marx guarantees its answer. But nothing in Marx forbids it either, and the first task of any political sequence is the diagnostic one. We are being harvested. The harvest is continuous. There is no outside. To see this clearly is already to have begun.