## I. INTRODUCTION In the tradition of, or in dialogue with, the political economy of communication (PEC), several authors have sought to understand the role of digital platforms for capitalist accumulation, mainly by proposing a new general paradigm that updates Marx's conception of the exploitation of labor through the extraction of surplus value. This theoretical tendency arises from the fact that the PEC seeks to analyze the processes through which society supplies itself with symbolic goods "under capitalist conditions of production and consumption, including its political and institutional processes, taking as a starting and ending point the labor theory of value" (DANTAS, 2012, p. 286). In this sense, Dantas (2012) proposes a general law of information capital; Bueno (2017) states that the attention economy is the central contemporary form of value production; Fuchs (2013) defines digital work as its source; and Bolano (2000, 2012) offers a general theoretical framework for analyzing the cultural industry and information commodities, including in the digital context. New "general laws" and "general theoretical frameworks" are certainly ambitious interpretations that, by proposing a single logic capable of synthesizing all the multiple economic relations traversed by the digital, constitute a field of controversy centered around two questions: whether and in what terms the category of "surplus value" remains valid to characterize all contemporary processes of labor exploitation; and if so, who are the exploited people from whom surplus value is extracted in digital platforms. For some authors, the digital economy does not produce value (DURAND, 2017); for others, the extraction of surplus value is reduced to the exploitation of formal employees of companies such as Alphabet (BOLANO, 2012), and in some cases extends to all users of these services, almost half of humanity (DANTAS, 2012; FUCHS, 2013, BUENO; 2016). In the next section, this article describes the central lines of argument from each of these perspectives, in order to then present a theoretical alternative for understanding the role of the Internet and digital platforms for capitalist accumulation. To this end, in the second section we recover Marx's theory of productive and unproductive labor, particularly his reflections on the labor that produces intangible goods. This is necessary in order to consider the applicability of these concepts in analyzing the diversity of social relations mediated by digital platforms that involve the production of content, data, and metadata. The third section examines the role of click farm workers and ordinary platform users in the digital economy. In this sense, we develop an argument for replacing the search for a single general paradigm of digital labor for all digitally mediated relations with an understanding that they constitute a field of multiple accumulation strategies by different capitalists. Thus, the controversy over whether or not the activities of professionals from Internet companies and ordinary users of the platforms should be generally recognized as productive labor gives rise to a case-by-case analysis of the role they play at different times in the private appropriation of socially produced wealth, considering the combination of processes of surplus value extraction with what Harvey (2005) calls accumulation by dispossession. In this sense, the contributions of authors such as Durand (2018), Bolaño (2012), Dantas (2012), Fuchs (2013), and Zuboff (2018) become complementary rather than contradictory. Finally, in the conclusion, we announce our next steps of research agenda, developing the debate of an "algorithmic prince", in the Gramscian sense of prince as conductor of hegemony, as a fundamental political condition for the maintenance and expansion of capitalist accumulation through digital platforms. ## II. THE DIGITAL ECONOMY AND THE PRODUCTION OF SURPLUS VALUE There is a tradition of Marxist thought, of which Machado (2017) is a typical example, that understands the processes of capitalist accumulation centrally through factory production, ignoring the possibility that the creation of intangible goods participates directly in the global production of value. An updated perspective on this tradition is represented by Durand (2018, p. 8, our translation): In the perspective chosen here, the exploitation of labor always plays a central role in the formation of a global mass of surplus value, but the focus is on the mechanisms of capturing capital (monopoly intellectual) that allows it to accumulate its profits, taken from this mass global surplus value, and limiting their direct involvement in exploitation (Foley 2013, 261). The digital economy is therefore a rent economy, not because information is the new source of value, but because controlling information becomes the best way to capture value. Like Harvey (2005) and other authors, Durand (2018) emphasizes that the diffusion of information technologies has allowed the fragmentation of value production chains on a global scale, with the distribution to the capitalist periphery of several stages of production previously concentrated in countries of central capitalism. The concentration of value and its return to the central countries then occurs through the tightening of intellectual property rights, which characterizes the monopolistic intellectual capital. This works through the decentralization of tangible assets, factories, inputs, for example, combined with the centralization of intangible assets. In this perspective, the role of digital platforms is to integrate and coordinate these global value chains, articulating automated production management and customer relationship systems. The control of infrastructures software offers a central role in the governance of production chains, which allows a disproportionate capture of value in exchange. In this way, although the production of value is mainly transferred to the periphery, the headquarters of the companies in the central countries accumulate value in the form of profits from the rent derived from intellectual rights and from the concentration of the benefits of network externalities in the intangible stages of accumulation. According to Durand (2018), this growing disconnection between the tangible and the intangible is accompanied by a powerful increase in the logic of value capture to the detriment of production, which contributes to contemporary stagnation and fuels the leap in financialization, which is also accelerated by the crescent digitalization and algorithmic automation of the stock market and other speculative investments. It is necessary to emphasize that there is a certain consensus in the PEC literature in recognizing the role that digital communication plays in what Marx (2011, p. 699) called "the nullification of space by time," that is, the acceleration of the circulation of capital through the development of communication and transportation. To accelerate the time of capital circulation, virtual platforms contribute as a counter-trend to the anarchy of capitalist production through the advertising function of communication (Bolano, 2000), which accounts for the majority of the revenues of giants such as Alphabet and Facebook (FUCHS, 2015). Today, digital platforms match supply and demand in an increasingly instantaneous and efficient manner, concentrating human attention on an unprecedented scale through the spectacularization of life, combined with user data extraction processes that allow algorithmic predictive effects and detailed market segmentation. In addition, platforms accelerate not only the exchange of information about the demand and production of material goods, but also the intangible nature of the production and exchange of cultural goods that occurs through them, allowing the expansion of the volume of transactions while shortening the interval between the consumer's decision and the fulfillment of his desire. Beyond this consensus, a group of authors (BOLANO, 2000, 2006; BOLANO, VIEIRA, 2014; BUENO, 2017; DANTAS, 2012, 2014; FUCHS, 2015) consider that immaterial production processes also contribute directly to the accumulation of capitalist production value through the creation of intangible goods. However, these thinkers disagree among themselves in defining which actors, whether platform users or their tech workers, would perform the productive work of transforming digital data into audience commodities, as will be developed below. Smythe (1977) was the first to systematize a theory of "audience work" that created the "audience" commodity. For Smythe, capitalism extended the productive working day into the domestic experience of individuals by transforming them, for example, into spectators: the spectators' time occupied by commercial media, during which their consumption tendencies are formed, would produce the audience commodity, sold in the form of advertising time in the broadcast media to those who would make them their consumers (SMYTHE, 1977). This contribution of Smythe is fundamental to the division of different authors in the analysis of the economic contribution of users and professionals of digital media to capital. In this sense, Bolano (2000) disagrees with Smythe (1977), as he considers that the production of the audience is not separated from the production of content by the workers of the cultural industry. Bolano (2000) carries out a broad review of the different PEC traditions in order to propose a general theory of the cultural industry based on the dual character of cultural goods. According to the author, "the work of professionals in the cultural industry would have the specificity of producing two goods at the same time: the object (program, newspaper, film) or cultural service and, at the same time, the audience" (BOLANO, 2000, p. 43). It would be the concrete activity of artists, journalists and producers, within the degree of relative creative freedom that limits their real subsumption under capital, that would produce the audience by capturing the attention of the spectators on the basis of their symbolic value. For example, the price and interest of the advertising market for a given audience would depend not only on the number of viewers, but also on the quality of the audience reached by the program through the "credibility," sophistication, or sensationalism conveyed by the professionals involved, qualities that advertisers may want or fear to be associated with their products (BOLANO, 2000). In more recent works, Bolano (2012, 2014) presents his hypothesis that, for the production of value on the Internet, the interaction of users and the content they produce do not have a directly productive dimension for capital. From this theoretical perspective, the attention and data generated by users, including the content generated by them, serve as input for the production of a commodity called audience through the combination of the dead work of algorithms with the live work of data analysts and programmers who are regular workers of the platform companies. The resulting audience is then sold by the platforms to third parties (BOLANO, 2012). Therefore, platforms like Facebook and Google only update a traditional business model of the cultural industry, in which the audience produced has the use value for other capitalists of guaranteeing potential sales, and the entire mass of surplus value comes from the work of its employees. Fuchs (2015) disagrees with Bolaño: for him, the information produced by users that allows the mapping of segmented audiences would not only be inputs processed by others, but the fruits of their own work as prosumers. According to Fuchs (2015), prosumers are the users whose consumption is immediately unpaid labor disguised as free access to the platforms. Fuchs (2015) and Dantas (2014) agree that the users' activities produce surplus value, although they disagree on how the extraction of surplus value defines the amount of value. The work of paid technicians on digital platforms also participates in the construction of data useful to advertisers, according to Fuchs (2015); however, this wage labor combines with the free labor of users in what Marx (1978) called socially combined labor. In socially combined labor, "the cooperative character of the labor process itself [...] necessarily extends the concept of productive labor" (MARX, 2013, p. 136). In order to "work productively, it is no longer necessary [...] to personally put one's hand to the work; it is enough to be an organ of the collective worker, performing one of its subfunctions" (MARX, 2013, p. 136). For Fuchs (2015), in this combination, the work of ordinary users is even more crucial than that of the company's workers, since the contribution of the latter is largely already frozen as dead work in the form of codes, algorithms, and automated procedures. In contrast, Fuchs argues that if ordinary Facebook users refuse to access and interact on the platform, the company immediately loses the ability to continue to provide its value to advertisers, that is, to sell advertising, since its users are a fundamental asset for its business model. Therefore, as the author details, the price of an ad on Facebook represents the value of the audience on these platforms, produced according to the average time that the segmented audience of users spends paying attention to Facebook, divided by the average number of ads presented to them in that period (FUCHS, 2015). For Fuchs (2015), this means that we have the production of a classic commodity. The value of advertising space and the attention of potential buyers are the result of the expenditure of users' working time, whose work has a concrete dimension - the specific information of each user - and an abstract dimension - the general audience time, which serves as a measure of value. And since this time is absolutely unpaid, the surplus value is extracted absolutely. Contrary to Fuchs (2015), Dantas (2012) claims that the productive work of users does not produce a new commodity, since the information they produce, like all information, has properties that make it difficult to transform it into a commodity. To reach this conclusion, Dantas (2012) compares the characteristics of tangible and digital goods in terms of their possibility of being subject to the commodity relationship. According to Dantas (2012), tangible goods are rival goods because when they are sold, their ownership is transferred to the new owner not only as a legal relationship, but also as a concrete possibility of consumption. Moreover, the exchange value of a tangible good can be measured in terms of the human labor time consumed in its production (MARX, 2013). However, information has the property of being reproducible, "consumed" by an infinite number of people at the same time; information is a non-rival good. Moreover, the reproduction of information occurs at a speed that is independent of human working time; its value is not related to the expenditure of abstract labor, although its production still depends on the concrete work of its creator (DANTAS, 2012). In conclusion, information is difficult to transform into a commodity because access to it is difficult to privatize; it is difficult to exercise absolute ownership over its availability in order to create scarcity, in addition to the fact that its production does not take place according to the law of value, which allows for the equalization of the exchange value of commodities in the capitalist system. In this context, in order to maintain private property over the information produced by users, in order to negotiate access to it in the market, platforms must use extra-economic coercion, such as intellectual property rights. For example, it becomes a crime to copy proprietary information. At the same time, in order to prevent the violation of intellectual property not only in law but also in practice, platforms try to monopolize the audience in environments where the copying of data and code is technically prevented through encryption, login systems and restrictive terms of use, what Dantas (2012) calls "walled gardens". Therefore, according to Dantas (2017), the added value extracted by the free work of users is not realized as profit from the sale of a commodity called audience. The profit comes from the monopoly rent that platforms earn by giving advertisers temporary access to the live activity of users' interactive audiences. As already mentioned, the value of this information is not related to socially necessary human labor time, but Dantas (2012, 2014, 2017) considers that the category of surplus value still applies to the process of capital accumulation on digital platforms, due to the fact that the attention and interaction time that users dedicate to digital platforms is not paid for. Like Fuchs (2015), Dantas (2012) argues that the exploitation by capital of users' life activity is a form of extraction of absolute surplus value, in which the extension of the part of the worker's unpaid journey has reached the point of your entire working time. However, to the extent that this surplus value occurs in the production of information that does not become a commodity, it is not realized in the form of profit. It must be accumulated from rights to wealth in the form of rent. Fuchs (2015) argues otherwise, stating that Marx's definition of rent is that of wealth obtained not as the result of human labor, but as property rights to goods such as land and nature. As interpreted by Harvey (2005), it is sufficient for the capitalist owner of land to maintain legal ownership over it and wait for a moment of scarcity to rent it, or to "produce" scarcity by claiming that his land has a unique character. Rent therefore means the consumption of surplus value and not its production, since the rentier owner appropriates part of the value produced by society. Therefore, according to Fuchs (2015), it is not possible to consider as productive work of value an activity whose contribution to the social production of wealth, in order to be privatized, depends on rent mechanisms. ## III. DIFFERENT STRATEGIES OF ACCUMULATION: SURPLUS VALUE, FREEDOM AND DISPOSSESSION So far, we have analyzed several concepts that have been proposed to understand globally how the Internet and digital mediation contribute to the accumulation of wealth under capitalism. Without endorsing any of them, this article proposes that the Internet is an ecosystem that is not traversed by a single mode of surplus extraction or production, but by different strategies through which actors privately appropriate socially produced wealth. These strategies will now be analyzed in order to consider, in each case, the adequacy and limits of the previous author's propositions. It is necessary to remember that Marx (2011), in proposing economic categories to understand social relations, takes into account the political determination of these relations, that is, the collective and individual interests of different actors that govern their actions and the correlation of forces between them. For example, in Marx's (1980) definition of unproductive and productive work, we have that productive work is that which is directly exchanged for capital, that is, that which is already subordinated to the logic of capitalist accumulation, in what Marx (1980, 2013) defines as subsumption. Therefore, "the specific character of productive labor is in no way linked to the concrete content of the labor" or to "the nature of its product" (MARX, 1980, pp. 137-128). The same kind of labor may be productive or unproductive. For example, when Milton wrote Paradise Lost for five pounds, he was an unproductive worker. In contrast, the writer who works for his editor in the manner of factory labor is a productive worker. Milton produced Paradise Lost for the same reason that the silkworm produces silk: by an impulse of his nature. Then he sold his work for five pounds. But the Leipzig intellectual proletariat who produces books (e.g., compendia on economics) under the direction of the publisher is a productive worker; for his product is subsumed under capital from the beginning, and only comes to light to increase its value. A singer who sells her singing at her own risk is an unproductive worker. But the same singer, if hired by a businessman to make money from her singing, is a productive worker, for she produces capital. (MARX, 1978, p.76, emphasis in original). From the point of view of the production of value, it is irrelevant whether the commodity produced is material or immaterial, tangible or intangible, and its use value "may be totally insignificant" (MARX, 1980, p. 138). What matters is the absence of freedom: the production of value means the hegemony of the logic of exchange value over use value, that is, the subordination of free creation, of the producer's interests and desires to the sole objective of accumulating his employer's capital. In this sense, the definition of value production proposed by Durand (2018), based on the tangible or intangible character of the goods produced, can be considered unsustainable in Marxist terms. However, Marx (1980) makes the warning that while these types of immaterial labor may contribute to the accumulation of the individual capitalist who employs them, they are insignificant to the production of value as a whole. This begs the question: wouldn't it be contradictory if productive labor, from the point of view of the capitalist who employs it, were insignificant from the global point of view of capital? After all, does this labor produce value or not? Marx adopts a perspectivist definition: [...] the fixed definitions of rent and capital exchange places with each other, appearing, from the point of view of the isolated capitalist, to be relative definitions that disappear when we consider the global process of production. [...] It is possible to overcome the difficulty if we imagine that what is rent for some is capital for others, and that these definitions therefore have nothing to do with the effective particularization of the components of the value of the commodity. (MARX, 1981, p. 969) To the extent that rent and capital are social relations, or categories that describe these relations, different subjects can simultaneously have different relations to the same portion of socially produced wealth. To a large extent, private appropriation occurs at the same time that wealth is produced under the control of capital, in what Marx (2013) called the production of value, which he explained in terms of the extraction of surplus value and which is commonly associated with a description of the industrial process of material goods. In these situations, it simultaneously contributes to increasing the general wealth of society and establishes private property over this part added to the whole. In other cases, the activities only produce private rights over the wealth produced by society as a whole, so the capitalist involved in them accumulates his capital by extracting money whose value originates from other sources, in what Marx (2013) calls rent. As a social relation, however, the definition of value is essentially political. A decision such as the privatization or socialization of the means of production, and in this case the choice between self-management or state control, has far greater historical economic effects on the accumulation of value at a given moment than the industrial production of an entire country. In this context, the applicability of the categories proposed by the authors discussed above for analyzing the social relations mediated by the Internet depends on how the producers, intermediaries and consumers of attention, content, data and metadata understand their own activities and the interests that guide them. ## IV. CLICK FARMS AND USER-GENERATED ### CONTENT: PRODUCTIVE LABOR AND #### DISPOSSESSION The general logic of the cultural industry described by Bolaño (2000) applies perfectly to click farms, social interaction factories, or the commercial production of disinformation, and also partially explains services such as Netflix, YouTube producers, and some digital influencers, as their workers recognize: "Of course I made money by publishing fake news, but Google made more," says Christian, 19, a young man from Macedonia who works in a disinformation factory (TARDÁGUILA, 2017). Christian is an employee whose sole goal in producing content for the Internet is to reach a measured audience through user interaction, for which his company receives a portion of the advertising revenue. The Macedonian tested the political positioning that generated the most clicks on the Internet: "Not Hillary, not even Bernie Sanders. Trump won" (TARDÁGUILA, 2017). It is therefore the example of an intellectual proletariat described by Marx (1980): his product is subsumed under capital and only comes to light to increase its value. However, the company where Christian works doesn't sell the content it produces; it is freely published online with the aim of generating an audience that is offered through Google's auction system. The company also does not acquire the data of the users who access its services, monopolized by Alphabet, which only provides some information about the audience to the owners of the sites. In this sense, the model of the cultural industry proposed by Bolano (2000), based on the duplicity of cultural commodities, is updated in this case: producers offer their content for free, expressing a real subsumption to capital, and at the same time they lose control over the offer of the audience they produce. One can assume, like Fuchs (2015) and Dantas (2014), that to the extent that advertising depends on user interaction, users also contribute to the process of audience production. This does not detract from the fact that salaried content producers for the Internet are perfectly included in Marx's (1980) analysis of productive workers of value; they even understand themselves as such, like Christian, who understands perfectly that most of the wealth resulting from his work does not remain with him. In the case of click farms, however, the interactions on social platforms, such as clicks or likes, are produced exclusively by salaried professionals. They are workers in miserable conditions, worthy of the descriptions in Marx's Capital: "[...] they sit in front of screens in dark rooms with windows covered by bars, sometimes working at night. To do so, they must generate 1,000 likes or follow 1,000 people on Twitter to earn a single US dollar" (ARTHUR, 2019). Arthur (2019) describes the interaction industry that combines precarious work in Bangladesh with a legal platform façade, crowdsourcing. Crowdsourcing is a means of social collaboration inspired by the logic of crowdfunding, collaborative financing, through which users can exchange goods or services among themselves without monetary intermediation, in a barter process. Crowdsourcing it is one of the practices of the "new economy", a showcase of a supportive, creative and cool capitalism, based on decentralized models and distributed exchanges, the latest version of "California ideology". Crowdsourcing platforms contribute to users sharing rides or practicing couchsurfing, the free accommodation of tourists in the homes of hosts who, in exchange, will one day stay in the homes of other users. The oligopolization of the Internet on a global scale, based on the algorithmic mediation of online content, has largely extended to the algorithmic mediation of crowdsourcing practices that originally emerged as nonprofits facilitated by digital technologies. Uber has commodified the provision of rides, just as Airbnb has built a business model inspired by the culture of couchsurfing. In the case of the crowdsourcing service Shareyt, analyzed by Arthur (2019), despite the appearance of being a service for the free exchange of likes between users, about 30 or 40 percent of the clicks came from Bangladesh factories. This is a reversal of the paradigm of Bueno (2017) and advocates of cognitive capitalism, in which capitalism accumulates wealth by tracking spontaneous human relationships that occur outside the disciplinary logic of work. In click factories, workers create interactions mechanically, completely alienated from their personal desires or interests, producing fictitious digital trails that simulate for their customers, brands, and digital influencers the capture of desire and attention from fake profiles, from a non-existent population. At the same time, the human nature of these factory workers is what makes them fool the filters of digital platforms that are capable of blocking automated interaction actions. Therefore, both in the salaried production of likes and in the production of misinformation, we consider that the accumulation model based on the extraction of surplus value remains valid, especially absolute surplus value due to the extension of the working day, precariousness and payment of wages of hunger. This is a fusion between Bolano's (2000) proposition of the production of surplus value by salaried professionals and that of the audience as an interaction produced by users proposed by Fuchs (2015) and Dantas (2014, 2017). However, much of the content, interactions, and digital traces produced on the Internet are not the result of paid labor, but of activities by platform users motivated by their own interests and perceived as consumption of services offered by Google, Facebook, and other companies. Can work be considered an activity that is not perceived by those who perform it? In particular, can we consider value-producing labor as an activity that develops on the basis of the users' own impulses and is not directly subordinated to the command and control of the capitalist, to the real subsumption of labor? The question may be who should answer this question. Considering the historical determination of the relations of exploitation not only as economic but also as political, the propositions of Dantas (2014, 2017) and Fuchs (2015) about digitally mediated social interaction as work may become valid to the extent that users themselves begin to recognize their activities as economically subordinated to capital, as an exploitation of their time, knowledge, and data, and begin to demand something in return beyond access to platforms. Bueno (2017) describes the debate on how claims of rights from users regarding their attention capacity equivalent to those of workers in relation to the sale of their labor power are already emerging: 1. Ownership: I own my attention and can store it safely privately; 2) Mobility: I can move my attention wherever I want, whenever I want; 3) Economy: I can pay attention to whoever I want and get paid for it; 4) Transparency: I can see how my attention is being used.(GOLDSTEIN, 2005 at GOOD, 2017, p. 56) Silveira (2017) points to an identical logic in relation to the data market, with the emergence of proposals that consider that data and metadata producers should be remunerated in exchange for their process of alienating rights and control over them. However, according to David Harvey (2005), there is an alternative understanding, originating in the work of Marx and developed by Rosa Luxemburg, that captures processes of wealth accumulation by capitalists without depending on the production of value through the extraction of surplus value. This is the first process of capitalist accumulation in history, which Marx (2013) called primitive and, when it occurs in a contemporary way, Harvey (2005) calls dispossession. In the context of the primitive accumulation of capital described by Marx (2013), Bolano (2000) highlights what he calls the "primitive accumulation of knowledge": to the extent that the knowledge previously exclusive to workers was appropriated by capital, along with scientific knowledge through intellectual property, the conditions were created for the incessant technical development of the productive forces in capitalism, a historical process also described by Dantas (2012). In this sense, Harvey (2005) revisits the concept of primitive accumulation, which he renames dispossession, when he describes how processes of mercantile accumulation occur simultaneously through the transformation of various forms of "property rights - common, collective, state, etc. - into exclusive property rights; [...] and the suppression of alternative forms of production and consumption, including resources" (HARVEY, 2005, p. 84). - and the suppression of alternative forms of production and consumption, including natural resources" (HARVEY, 2005, p. 84). The advantage of the category of dispossession applied to the production of attention, interaction, and data is that instead of proposing the regularization of this activity appropriated by capital as waged labor, that is, the formal recognition of exploitation, it offers a non-market alternative for understanding the fruits of these activities: the common good, or what Marx (2013) called the commons. In short, we proposed that the interactions between users and their results on the Internet be understood as a kind of common good, a wealth produced by humanity as a whole, but which is immediately dispossessed by proprietary platforms. Messages, photographs, knowledge and content in general, produced by users on the basis of their use-values and previously governed by non-commercial principles, become private property under the mercantile logic of corporations once the terms of use of their social platforms are accepted, which undermine or destroy universal rights such as privacy and confidentiality. The private appropriation of wealth occurs at the moment of its production, but not because its production has been subsumed by capital, but rather through the extra-economic legal coercion of terms of use or through the oligopolization of the mediation of attention flows and Internet connection. An example is the appropriation by the Google algorithm, in its search engine and word auction system, of content from indexed non-commercial sites such as Wikipedia or pirate sites. In this way, the current moment does not mark the epistemological exhaustion of Marx's theory of value (2011, 2013), which would be incapable of grasping new processes of valorization based on immaterial work, as Bueno (2017) argues. It marks the concrete exhaustion of wealth accumulation through the appropriation of other people's working time, which becomes, according to Marx (2011), a miserable measure for the potential for wealth production, in what Bensaid (2013) calls the miserability of the value. This is because the moment of General Intellect is not just one in which socially produced knowledge becomes available to everyone, as advocated by Marx and Engels(1961), but also one in which it becomes incorporated in the form of machines and automated processes increasingly autonomous in relation to humans (MARX, 2011). It is the peak of the organic composition of capital, of the successive replacement of humanity's living labor by the dead labor of machines, of reification: the moment of humanity's general intellect converted into productive force. The miserability of value is the anticipation by Marx (2011, p. 943) that, as automation becomes potentially universal, it no longer makes sense to base an economic and social system on the exploitation of employees and mass wage labor, because the means of satisfying needs through social cooperation between men and machines become abundant, calling into question private property and the private accumulation of wealth as a social logic. Dispossession, unlike the analyzed propositions centered on new modes of value production, considers that, even if the wealth extracted from free activity appears in the form of surplus value for the capitalists who own the platforms, it appears as rent from the global point of view of capitalism. This is the difference in relation to Bueno (2017), for whom capitalism can continue its permanent expansion based on new immaterial sources of value. However, in opposition to what Dantas (2014, 2017) defends, it is not about rent arising from work subsumed under capital, but rather the dispossession of free activity constitutive of a common good, the General Intellect. The accumulation of data, interactions and digital content as a global expansion of surplus value is hampered not only by the particular properties of information as a "commodity" (DANTAS, 2014, 2017), but also by the inherent difficulty in trying to grasp the wealth produced by General Intellect in the "miserable form of the theft of working time" (MARX, 2011, p. 943). According to Marx (2011), the emergence of General Intellect marks the moment in which the private appropriation of humanity's production by a small portion of it enters into a profound contradiction and dominant interests can only remain as barriers to free creation. Therefore, cognitive capitalism and the enormous wealth accumulated by internet oligopolies are unable to reverse the continuous and accelerated fall in the profit rate, demonstrated using different analysis methods by Toshio (2017). ## V. CONCLUSION The perspective that seeks to define all digitally mediated activity as labor truly subsumed under capital, and to demand remuneration for it, is to formalize and at the same time legitimize a new form of exploitation. Meanwhile, dispossession shows that capital does not take possession of most of the wealth produced on the Internet by fulfilling a productive historical role, but in a violent way through legal coercion and the violation and destruction of rights. This also means that capital, through technological development, has not inaugurated a new era of expansion of its accumulation through new productive processes of value, but that it can only continue to exist in increasingly fictitious forms based on the extraction of rents. There is an enormous production of wealth in the form of new relationships and products capable of satisfying immaterial human needs, of "fantasy" as defined by Marx (2013), but since these tend to become common goods and capital can only appropriate them in a coercive way, the resulting accumulation is only monetary and derives from the ability of Internet corporations to capture investment in the financial market and rent in the advertising market. Therefore, the accumulation of capital through digital platforms is a symptom of the fact that the domination of capital increasingly depends not on the economic efficiency of its mode of production, but on its violent domination of human life, exercised through other relations of power that allow it to continue to focus attention and exploit data. This is a contradiction that, as Bolano (2008) concludes, updates the meaning of the maxim "socialism or barbarism" and points to the urgency of a political solution. In this sense, as a future research agenda, we propose to analyze how accumulation relations mediated by digital algorithms on oligopolistic platforms are legitimized by the political effects of algorithmic mediations on these same platforms, and what counterhegemonic strategies to these effects would be. For development in subsequent work, we argue that the category of "electronic prince" created by Ianni (1999) to characterize broadcasting agents can be updated to "algorithmic prince," a new politically dominant capitalist fraction organizing the new power relations and accumulation that have emerged with the digital. Specifically, the algorithmic mediation of an increasing proportion of human relations under the control and ownership of digital oligopolies suggests that there has been a process of change not only in the quantitative but also in the qualitative conditions of hegemony, an intrinsic dimension of the capital accumulation system. Declaration of Conflicting Interest The Author declares that there is no conflict of interest. Funding This study was financed in part by the Coordenação de Aperfeçoamento de Pessoal de Nível Superior - Brazil (CAPES) - Finance Code 001

Generating HTML Viewer...