Factories or workhouses? Digital platforms and the re-feudalisation thesis

Re-feudalisation

As Gane (2024) has stressed in his critical review of the techno-feudalism thesis, periods in which rent seeking and the erection of large monopolies prevail over profits and market competition are a recurrent feature of the history of capitalism. And he is not alone: scholars such as Saito and Sasaki (2025) and Gilbert (2024) recognise the rising importance of rent in contemporary capitalism, but they both argue that this, in itself, is not enough to declare a new mode of production. Gilbert (2024) cites other theorists, such as Piketty and Goldhammer (2014) and ‘neo-workerist’ economists such as Fumagalli (2019) and Vercellone (2013), who analyse the increase in wealth derived from asset ownership (including rent) without concluding that capitalism has been supplanted. The ‘becoming rent of profit’ (Vercellone, 2013) is an inherent tendency throughout the history of capitalism, not something that in itself brings about its demise.

Many scholars from Marxist and post-Marxist traditions have also contested the refeudalisation thesis and criticised its conceptual confusion. Morozov (2022) criticised the technofeudalism argument for resting on an overly dualistic conception of capitalism. Gilbert (2024) argues that we are confronted with the rise of ‘platform capitalism’ as a new capitalist ‘regime of accumulation’ – a new configuration of technologies, institutions and socio-cultural practices. Similarly, Saito and Sasaki (2025: 12) advance the concept of ‘rentier capitalism’ (see also Sadowski, 2020), where the source of rent lies ‘in the continual enclosure and dispossession of the commons’ and argue that this constitutes the most advanced and contradictory form of contemporary capitalism. They suggest that ‘to move beyond the conceptual confusion surrounding technofeudalism, it is useful to return to Marx's Capital, especially his theory of rent (2025: 2).

In other words, that these features have come to dominate digital capitalism does not necessarily mean that we are on our way to ‘something worse’ – the re-introduction of some imaginary feudalism where ‘economic life did not involve economic choices’ (Varoufakis, 2023: 19) – but rather that capitalism is confronted with the limits to expansion provided by the very social structure that it has erected. Indeed, in his reading of Marx (and Braudel), Arrighi (2010) shows that the systemic dominance of rent over profit is a feature of what he calls the ‘autumn phase’ of a cycle of capitalist accumulation. Capital's ability to further develop the forces of production declines: an old idea of how to make money and grow markets runs out of steam and capital tends to divert to financial markets where it looks for flexible short-term speculative opportunities as well as for long-term strategies of monopolistic rent seeking. As Marx characterised the basis of every potentially revolutionary situation in the preface to his Contribution to the Critique of Political Economy: ‘the forces of production have become too advanced for the prevailing relations of production’ (Marx, 1977 [1859]: ii), which now, block rather than support the former's full development. In other words, re-feudalisation is also, as Dean (2025) points out, a crisis of the imaginary: marked by a material and cognitive inability to imagine a different kind of future.

This was precisely what Fernand Braudel (1974: 2229–31) saw in his analysis of the decline of the Kingdom of Naples in the seventeenth century, where, to our knowledge, the idea of a return of feudal relations of production was first invoked. Faced with a slowing economic growth resulting from the shift of the centre of economic dynamism to North Western Europe, merchant capitalists invested in landed estates and, importantly, tax prebenda, which the Spanish Viceroys, faced on their part with the task of financing their continuous wars, were happy to sell. This caused a strengthening of feudal structures in the countryside, and an intensification of exploitation, along with a growing tax pressure. The rent thus extracted went increasingly towards the financing of conspicuous consumption on the part of the elites. It also generated a growing surplus population in the form of a destitute urban plebs – refugees from renewed forms of feudal exploitation in the countryside. As Braudel emphasises, the movement towards capitalist modernisation was temporarily stalled by these developments. Similar forms of re-feudalisation marked many areas of Europe during the ‘general crisis’ of the seventeenth century – parts of Eastern Europe, for example, saw the re-introduction of serfhood as a response to peasant rebellions and the expansion of the market for grain exports. In England (and the Netherlands), this crisis was overcome by enacting a fundamental transformation of the social structure of the ancient regime. During the 17^th and 18^th centuries, the feudal and agrarian nature of English society was dismantled to pave the way for the development of modern industrial capitalism. This Great Transformation happened ‘from above’, as the power of the merchant and land-owning bourgeoisie had been consolidated in the English revolution; and it also happened from below, as the ancient plebeian society with its notion of moral economies and traditional entitlements was dismantled in favour of a new ethos marked by industriousness and a market mentality. The Kingdom of Naples, along with other parts of the now declining Mediterranean, instead remained stuck, more or less, until the aftermath of the Second World War. Lesson: re-feudalisation is a recurrent feature of capitalist development. It sets in when existing social relations have become an obstacle to further development of the forces of production. It can be overcome only by a fundamental transformation of the structure of those social relations themselves. By finding a new way to ‘fix’ the remaining spatial and temporal ‘fixes’ that now block development (Harvey, 2001). During the 18^th and 19^th centuries, workhouses were essential to achieving such a transformation.

Workhouses

Workhouses became a common institution across Europe in the eighteenth century. In England, the erection of municipal workhouses continued a tradition of parish-based poor relief that went back to Elizabethan times, and their institutional form built on experiments with confinement and isolation as attempts to contrast the recurring plague epidemics that marked Europe from the fourteenth century onwards (Hindle, 2004). In the eighteenth century, the growth of pauperism – in the form of an urban plebs resulting from the collapse of agrarian society that followed the commercialisation of agriculture and related enclosures – seemed to overwhelm the capacity of the older parish-based poor relief system. Workhouses now became central to what Foucault (1975) called ‘the Great Confinement’, society-wide attempts to confine the novel surplus population – paupers, vagrants, lawless plebeians – into novel institutions that could provide discipline and reform. This way workhouses were only marginally concerned with extracting value from labour. While some did operate commercially – taking on often menial production tasks such as stone breaking, grinding bones for fertiliser or recycling rope fibres – labour was for the most understood as a disciplinary and educational activity, rather than as something to be organised around the demands of market profitability (Dunkley, 1981). During most of the eighteenth century, workhouses were still ideologically part of an ancien régime society, which they would eventually contribute to sweeping away. Their disciplinary purpose was not yet, or at least not only, the formation of a market-oriented homo economicus, but rather the re-integration of paupers within the moral confines of Christianity and, eventually, their salvation. As read the motivations for the construction of the poor house (Albergo dei poveri) in Naples in the 1750s: ‘Beggars were lost souls in need of redemption, because theirs was a profoundly immoral life in which sin was born from idleness: they cohabitated outside of marriage, cursed, engaged in blasphemy, gambled, drank, and died on the street like animals, without the benefit of the sacraments of the Church’ (Pallan, 1995: 10, our translation). Here, work was principally understood as a source of redemption, a quasi-religious exercise. As Karl Polanyi (1944) suggests, it was only in early nineteenth-century England, with the repeal of the Speenhamland legislation which offered a sort of guaranteed minimal income, that the workhouse came to fully assume its function in support of the market – as a sort of deterrent, offering conditions far worse than what could be obtained by labour market participation, in order to dissuade those not yet properly motivated from conducting a life of idleness.

Workhouses were a central institution in the bio-political reform that became a precondition for the transition to industrial capitalism: the eradication of plebeian solidarities, the end to a culture of rights and entitlements, and in general the transformation of the ‘surplus’ population into a viable industrial proletariat equipped with a utility maximising mentality capable of valuing the drudgery of labour in terms of the pleasures of consumption and the pride of achievement. In this sense workhouses, along with vagrancy laws, the regulation of begging (including the badging of paupers) and the repression of drinking, ‘bawd music’ and popular festivals, accompanied the stick of factory discipline and the carrot of consumer culture in accomplishing the Great Transformation that enabled England to exit the crisis of the seventeenth century as the pioneer of a new industrial society. Indeed the development of workhouses was by no small amount triggered by movements of plebeian resistance: from the wave of riots that shook Europe in the seventeenth century (at the time symbolised by the Masaniello rebellion in Naples in 1647 – Villari, 1985) to the early nineteenth-century Swing riots where agricultural labourers in Southern England protested against mechanisation and the assertion of a market regime (Hobsbawm and Rudé, 2014).

Throughout the transition to industrial capitalism, workhouses prepared the way for the internalisation of production in the form of industrial capitalism, where commodity production had been subsumed under the capitalist value form. Their role was particularly important in the first half of the nineteenth century as the mechanisation of agriculture created a growing rural surplus population that could be transformed into an industrial proletariat. Workhouses contributed to this transformation: Together with the reform of the poor relief system, they helped enforce labour market participation (by acting as deterrents from idleness). At the same time, they took part in prefiguring new productive relations and developing a novel ‘industrious’ subjectivity based on the linear relation between time and value. Can we see platforms as having a similar historical role today, in relation to the internalisation of reproduction?

Platforms as (digital) workhouses

In the last two decades or so, platforms have inserted themselves in all walks of life: production, consumption, distribution and sociality. One important condition for their success has been the availability of a massive ‘surplus’ population available to work for the meagre wages that they provide. Bonini and Treré (2024) have argued that contemporary climate and geopolitical crises provide the gig economy and the AI industry with an ever-growing reserve army of labour. A globally dispersed complex of refugees, slum dwellers and victims of occupations is compelled through immiseration, or law, to power the machine learning of companies such as Google, Facebook and Amazon. In this situation, platform labour has become an alternative to traditional forms of employment for migrants, the unemployed and the marginalised. As Milagros Miceli suggests in a documentary about content moderation platforms like Remotask, they employ the ‘otherwise unemployable’ (DW Documentary, 2025, cf. Kapania et al. 2026): the surplus population created by the combination of a collapse of traditional life forms (like the agrarian peasant economies undermined by a combination of agribusiness monopolies, land grabbing and climate change) on the one hand, and the contraction of the global market for old style stable ‘Fordist’ jobs, on the other.

Riders, Amazon workers, content moderators and click/data-workers in general are mostly drawn from the migrant population (Miceli and Posada, 2022; Cant et al. 2024, van Doorn, 2017), the under- or unemployed. Platforms transform them into an underpaid source of servile labour employed to provide even the middle classes with the comfortable experience of having servants that can be commandeered for leisure tasks regardless of economic rationality – delivering pizzas; maintaining digital spaces as bland and safe; and training AI to generate novel forms of cultural slop with greater reliability. This way, platforms have absorbed and put to work parts of the surplus population generated by the collapse of the Fordist industrial economy: ‘collapse’ both in the sense of automation-driven decline in employment and in the sense of the disintegration of the world-system hierarchies that, until recently, maintained most of that surplus population in the Global South where it would supposedly be absorbed by future industrialisation and development.

'Surplus’ population is not an absolute but a relative term. It is true that the world population has grown significantly over the last decades, but the source of the growing surplus population is rather the diminishing need for old-style labour power in the reproduction of capital, in part driven by the twin tendencies of platformisation and automation themselves. The definition of the contemporary ‘surplus population’ would then be ‘those excluded both from the wage relation and from stable sources of financial rent’. Like the early modern plebs, they do not participate directly in capital accumulation and consequently persist in a general state of pauperism and precarity. They might be instrumental to the extended process of accumulation by forming a reserve army of labour that contributes to keeping wages low and that can be mobilised in times of crisis or need. But this presupposes that a novel form of capital accumulation – such as industrial capitalism and the factory system – has become hegemonic and that there are few alternatives. The early modern plebs was not yet a functional reserve army, indeed they were understood as an obstacle to capitalist development that needed to be done away with, whether internalised as an industrial workforce, or simply removed: deportation to the colonies was a common way to deal with the plebs in eighteenth-century England, as well as in the Spanish empire (Mawson, 2013, cf. Ruda, 2011). Today, the position of the global surplus population is similarly ambiguous: Novel plebeian economies, such as the bazaar economy, crypto trading and informal or illegal trades are growing in importance and provide possibilities to survive outside of the wage relation or in ways that have but indirect (if important) connections to the official financial economy. The need to repress such potential plebeian autonomy has moved to the heart of security-oriented policies, in particular in relation to migrants. Platform labour is then one important way in which the surplus population can be absorbed and internalised as a flexible reserve army.

One aspect of this is providing an identity and a sense of purpose. This way platform labour might be comparable to the ‘bullshit jobs’ (Graeber, 2018) that keep the middle classes on corporate payrolls – (For now! Such ‘Bullshit jobs’ are among the most likely to be platformised within the context of a new wave of AI-centred automation.) Occupations such as HR, brand management or consulting, for which the direct productive contribution is unclear, provide a however shallow sense of meaning and subjective identity along with a justification for receiving a wage. In similar ways, some forms of platform labour might have an ideological, more than a directly productive function (albeit without providing a middle-class wage). For the contemporary surplus population, they provide marginal incomes that are difficult to make a living from alone, without the support of communal solidarities (Bonini and Treré, 2024) or a second (or third) job, and that do not generate enough surplus value to make platforms profitable in the old-style Fordist sense. Instead, like workhouses, they at least provide something in material terms, along with an identity as an industrious ‘deserving poor’ rather than an idle and morally corrupt pauper. (As the Financial Times suggested already a decade ago, Parisian Uber drivers disproportionally come from the Banlieus, and for them, Uber provides a route to social respectability – Chassany, 2016) In this sense, platforms might be more similar to the eighteenth-century philanthropically driven workhouses that provided fairly generous diets along with a lenient disciplinary regime, than to their nineteenth-century incarnation as ‘deterrents and ‘worst options’. Platforms do for the surplus population what bullshit jobs do for the middle classes: they absorb and include those not directly contributing to capital accumulation within the fold of neoliberal society, materially, and perhaps most importantly, ideologically. This makes them into components of a flexible reserve army that can be mobilised in case of need.

Ideologically, platforms also do something deeper and more revolutionary than simply imbuing an ‘industrious’ ethic (Arvidsson, 2019) in people who might otherwise have remained ‘idle’. Along with paid forms of platform labour, there is also the platform ‘free labour’ (if we want to call it that) of consuming, liking and sharing content, of playing and socialising (Gandini, 2021). While the platform economy has blurred the boundaries between ‘production’ and ‘consumption’ – as in the growing influencer economy or the rise of various digital hustles such as drop shipping or affiliate marketing – it does seem that the very design of platforms is oriented towards the absorption, not only of the surplus population as ‘industrious paupers’ but of surplus time in general. Whatever time is not deployed in the direct reproduction of capital should ideally be spent on the platform. Social media platforms updated, optimised and intensified the time-consuming and ‘cronophagic’ nature of traditional commercial media. Consumption time represents the final frontier of Netflix's economic valuation model, for example (Bonini, 2024). Audio and video streaming platform profits are intrinsically linked to their model of capturing human attention, their ability to colonise, maintain and increase users’ ‘dwell time’ on the platform (Seaver, 2022). Here, the ideological function of platforms reveals itself to be more far-reaching.

First of all, continuous interaction with platforms normalises a total datafication of life. Proponents of online privacy fight a losing battle against the fact that most people have already accepted that almost everything that there is to know about them – from their friendships and political views to their sexual kinks – has already been transformed into data that is used against them or ‘to improve their user experience’. This situation, the stuff of dystopian sci-fi only a couple of decades ago, has already been naturalised as a fact of life.

A second step in this process, the realisation of which we are seeing with the increasing role of algorithms and AI, is the institutionalisation of what Schwartz (2025) calls a post-choice society. This is a condition in which we are becoming habituated to the fact that our choices – political, consumer, sexual – are only partially our own, but largely pre-structured by algorithmically processed data droves. Together, these two principles – acceptance of the complete datafication of life and the deference of choice to recommendation algorithms and various ‘agentic’ AIs – constitute the basis of the rather dystopian ideal of a planetary supercomputer guiding our actions, which Benjamin Bratton (2022) suggested as a solution to safeguarding our survival in the Anthropocene. Given the transposition of the plane of civic action from intersubjective relations to the molecular composition of the ‘global biocommons’, the only ethical course of action is to defer our choices to the machine, submit and renounce our individuality for the status as ‘dividuals’ – datapoints in an algorithmically recomposed social beyond our political grasp.

Bratton's book was written during the Covid pandemic, some 5 years ago – a period obsessed with the issue of ‘security’. And what platforms might seek to achieve is the securitisation of the social (a tendency that is central to contemporary financial capitalism overall, and not just to platforms – Bryan and Rafferty, 2005). By this, we do not simply mean that they strive to render social relations ‘safe’. This is certainly an aspect of it. Platforms invest a lot of resources in censoring content and distributing it in ways that enhance the perception of ‘safety’ on the part of users. And this logic is now extended beyond the platform economy proper. As the director of US Immigration and Customs Enforcement envisions the future of the US deportation system to look like: ‘It needs to be like Amazon, trying to get your Prime delivery within 24 h’ but applied to human beings’ (Rodgers et al., 2025). In other words, extending the same platform logic to the deportation, persecution and internment of illegal migrants and, in the near future, probably other ‘unsafe’ elements of society as well (dissidents, activists and so on – cf. Kocher, 2025).

But securitisation also operates in the financial sense. Here, it refers to the transformation of insecurity into calculable assets. This seems to be the most important economic function of platforms, and this is perhaps the basis of the neo-feudal element of platform capitalism. Money is to be made, not primarily through the exploitation of salaried labour but through the transformation of the insecure – as in unpredictable – effervescence of human practice in diverse fields, such as love, politics and labour, into spaces where algorithms and datafication have extracted calculable risks that can be sold on financial markets as derivative assets. We are already seeing this at work in the massive proliferation of social and credit scoring mechanisms that have marked the last two decades of platformisation. And this might be the true source of the growing importance of ‘neo feudal’ monopolies: not the classic monopoly rent that comes from the power to command historically invested labour but rather a kind of speculative rent that derives from a monopoly over the ability to predict the future of a particular area of practice.

Rendering actions predictable – providing norms, roles and values that structure action in a stable way – is, of course, the basic function of social reproduction. Platforms perform this not through communitarian interaction or more generally ‘language’ but by decomposing actions into data points that can be recomposed algorithmically as risk estimates directed at financial markets. Here, we can draw an additional historical comparison. Workhouses prefigured and prepared for industrial capitalism by fostering a link between time and value: by enabling the labour theory of value to become an operational force rooted in social relations and subjectivity. Platforms are prefiguring a form of internalisation of reproduction via its securitisation: by transforming the ‘ordinary chaos of love’ (or life – cf. Beck and Beck-Gernsheim, 1990) into predictable relations between data points. This transformation of insecurity into risk (Knight, 1921) enables the transformation of social reproduction into financial assets. However, this also requires the promotion of a novel kind of subjectivity that is amenable to financial securitisation.

Indeed, the other side to economic securitisation is subjective. Here, platforms play an important part in promoting a subjectivity that dissociates economically gainful tasks from identity: the worker is no longer a carpenter, welder or plumber, she is someone who bids for and performs dissociated ‘tasks’ on Taskrabbit. The desiring subject is no longer a Krafft-Ebingian ‘pervert’ – defined by the singularity of focus that marks his sexuality – but an OnlyFans profile who engages in the kinds of ‘practices’ that clients bid for. This machinic decomposition of subjectivity – productive as well as leisurely – into discrete tasks is in many ways a culmination of that rise of algorithmic mediation that has marked modernity as a whole. Now, however, it is accompanied by an important ideological novelty. Platforms teach their users to accept and embrace the fact that the relation between tasks and renumeration is no longer linear. A glimpse at the feet of Valentina Nappi, the famous Italian porn star / intellectual, is worth substantially more than what ordinary OnlyFancers would charge. But this is not simply a matter of the Aristotelian theory of value – where the ‘just price’ should reflect the worth (or virtue) of the producer – transposed into contemporary celebrity. As Rosamond (2020) suggests, such celebrity is itself becoming increasingly volatile. Platforms, in particular the most recent ‘generation’ of increasingly ‘algorithmic’ media like TikTok (Gerbaudo, 2024), are habituating their users to accepting such volatility as a natural fact of life (Arvidsson et al., 2025). What is more, abandoning the linearity of rationality, rules or principles, and devoting oneself to the practice of volatility, habituating oneself to leave one attachment for another, to latch on to trending memes and trends, to pump and dump, might be the most valuable trait of platform labour: an ability to generate constant volatility that requires algorithmic securitisation in order to be packaged and sold as a prebendal source of rent extraction (cf. Thibault-Rochefort and McCurdy, 2025). The speculative subjectivity thus imbued in platform workers – promoted by what Petrosino and Volpe (2026) call ‘augmented precarity’ – thus contributes to generating further volatility, and as a consequence, increasing the value of algorithmic predictability.

Contrary to early modern workhouses, platforms perform this biopolitical function without much of an explicit ideology: there are few sermons, not much moralising and little of an overall social vision (apart perhaps from the occasional Silicon Valley pundit). Instead, the power of platforms is ‘ontological’ rather than ‘ideological’ to quote Lash's (2007) classic suggestion: they work by generating a reality where only certain actions and orientations make sense. Digital platforms are designed to ‘moralise’ users, educating them to internalise and accept a novel, emerging – if as of yet unclear – ethic. This process of moralisation occurs through the interaction design choices of the platform apps and the linguistic choices contained in the Terms of Service of these apps. These affordances function as materialised moral prescriptions, as Bonini and Treré argue (forthcoming): they seek to guide, discipline and shape users towards what designers deem the ‘correct’ use of the technology. In doing so, they aim to produce not just functional users but also morally aligned ones. In this sense, they function as ‘technologies of the self’ (Foucault, 1988), shaping workers’ identities by rewarding competitiveness and compliance. The enforcement of this ethos is delegated to algorithms and app interfaces, which embed the companies’ pre-scriptions (Akrich, 1992) into everyday digital labourers’ routines.

Platforms use technological design to moralise behaviour silently, crafting interfaces that nudge workers towards desired behaviours and operate through a mix of incentives and penalties. Gamification features and performance-based bonuses reward desired behaviours, while sanctions such as deplatforming or temporary account suspensions punish deviations.

Significantly, this moralising function is performed directly by private capital through the mediation of digital affordances, and not, as in the case of workhouses, via the mediation of ‘ideological state apparatuses’ such as the parish or the corporation of the poor. This way platforms can be understood as part of the direct biopolitical intervention performed by contemporary capitalism, what Fumagalli (2019) calls a ‘bioeconomy’. Central to this is the challenge of the overall internalisation of reproduction.

The internalisation of reproduction?

Workhouses were essential to what Arrighi (2010) called the ‘internalisation of production’. By this, he meant the direct commodification of productive activities as wage labour, and in general their subsumption under the capitalist law of value. This enabled the transition from a merchant capitalism where profits depended on the extra-economic command of external productive resources, to an industrial capitalism where – as Marx stressed – profits mainly derived from ‘the abode of production’ itself. It transformed production from an external source of wealth into a direct component of surplus value, thus radically expanding the circuit of capital accumulation and ‘fixing’ the ‘crisis of the seventeenth century’ by enabling the transition to what we now recognise as capitalism proper. Might platforms seek to achieve something similar with regards to reproduction?

This suggestion requires some unpacking: First, and perhaps most obviously, most of the activities that are presently being ‘platformised’ relate to the sphere of reproduction. True, the architecture and institutional logic of platforms themselves evolved from the novel requirements of outsourced – and later globalised – industrial production. Chiefly, they responded to the need to coordinate the complex productive networks that resulted from Toyotist outsourcing, and later global supply chains in the ‘80s and ‘90s (Terranova, 2022). They did this, however, by datafying the increasingly important reproductive aspects of production: achieving what Christian Marazzi terms the direct inclusion of ‘language’ as an internal component of the production process in the form of self-managed teamwork, continuous quality control, knowledge sharing and the ‘collaborative communities’ that now were increasingly important to knowledge work, along with, importantly, logistic coordination (Marazzi, 2011, Adler and Heckscher, 2006). The implementation of platforms outside of the factory system and in social life that has marked the last two decades has predominantly targeted similar reproductive activities – communication, socialisation, affective fulfilment and identity formation, and logistics. Indeed, Jarrett (2015) has suggested that ‘platform labour’ is better modelled on the theories of reproductive labour that emerged from ‘70s Marxist feminism.

There is however a fundamental difference. In Leopoldina Fortunati's (1981) pioneering work, reproductive labour – in her case, the unpaid labour of the housewife – produces value by reproducing (mostly male) labour power. Household labour is not directly commodified. Rather, it generates value in what she calls a ‘multiphasic’ way. Reproductive labour produces the use values that, in turn, make possible the commodification of wage labour on the job market. Fortunati suggests that there is a ‘hidden unpaid contract’ between capital and women. In ways similar to what Moore (2017) calls ‘cheap nature’, reproductive labour constitutes a hidden resource that supports and enables commodification processes that occur at a different level. Contemporary platformisation, however, aims at the direct commodification of reproduction, without the mediation of family, community or other non-capitalist institutions. It aims at accomplishing this through extensive datafication that, in turn, allows for financial securitisation.

As in the case of the internalisation of production in the transition to industrial capitalism, this requires the overall transformation of attitudes and dispositions, the production of what Foucault (1975) called ‘docile bodies’ ready to submit to factory discipline. However, this re-composition of subjectivity is itself not possible unless there is a disintegration of an earlier social order: the overall ‘crisis of the seventeenth century’ in the case of early modern workhouses and the collapse of Fordist organised modernity in our case. Indeed, the raw material of platform ‘labour’ is the generalised ‘participatory culture’ that emerged as part of what Virno (2003) called mass intellectuality. The movements of the ‘70s strove to free subjectivity and, more generally, ‘language’ from the Fordist shackles of family, state and territory to enable the unstructured ‘creative’ experimentation with life forms as a mass phenomenon. In addition, from the ‘70s and on, the crisis of Fordist modernity and the neoliberal response that it triggered has created a growing surplus population of the un- or underemployed, the precarious and increasingly the downright destitute. Together with novel flows of migration, this has generated a neo-plebeian surplus population dwelling in the ‘surrounds’ (Simone, 2022) of financialised capitalism, who find their identity and livelihood outside of the traditions and institutions of a now half-forgotten Fordist modernity.

Starting in the mid ‘90s, the web enabled these, or at least the more privileged component that had access to internet connectivity and possessed the concomitant skills, to find novel, communitarian ways to organise their symbolic and social re-production. This constituted a significant empowerment and led to an explosion of identities, sexualities and forms of productive relations. Indeed, many intellectuals at the time envisioned a peer-to-peer civilisation based on the novel empowered forms of social reproduction that came with digital media (Bauwens, 2009). However, the incorporation of the Web into the fold of financial capitalism in the form of early ‘00 ‘Web 2.0’ began to transform this mass intellectuality into more structured forms of ‘interactivity’ (Jarrett, 2008) or ‘participation’ (Coté and Pybus, 2007) that promoted a subjectivity fit for the immaterial labour of producing commercially viable online advertising space. Initially, this internalisation of reproduction proceeded through the subtle promotion of an entrepreneurial self-dedicated to creative forms of self-branding, fit for selling oneself in neoliberal times (Gershon, 2014). After the financial crisis of 2008, such subjective measures were accompanied by objective affordances: the consolidation of simple classificatory practices such as hashtags or ‘like’ buttons (Gerlitz and Helmond, 2013; Fourcade and Healy, 2024), the growing algorithmic management of sociality and informational flows, the incorporation of logistics in the form of ‘delivery’ and ‘ride’ platforms and, in particular after Covid, the increasing automation of these processes. With contemporary developments of AI we have reached a point where a growing number of reproductive processes are automated: as in the case of the formation of consumer demand through recommendation algorithms and the targeting of advertisements, the substitution of AI for menial bureaucratic tasks, the introduction of chat bots in the form of friends or intimate partners and the growing presence of driver-less vehicles such as taxis and delivery drones (not to speak of military deployments where AI now also supplies the coordinating function of intelligence analysis and target acquisition). However, such automated mass intellectuality builds on the continuous participatory labour of the many (or, as Pasquinelli (2023: 12) put it, the ‘social intelligence that informs and empower AI’), albeit often in forms that have been reduced to repetitive clickwork, re-posting or scrolling. Like the de-skilling that stood at the basis of Taylorist scientific management and that began with the labour discipline imposed by early modern workhouses, platforms thus strive to reduce the reproductive energy of mass intellectuality to the structured and computable labour of a number of simplified, repetitive tasks that provide the raw material for algorithmic automation.

Seen this way, platforms could be preparing the way for the subsumption of reproduction as a direct source of value. As Arrighi stresses, a similar operation is a necessary step for any kind of resolution to the current crisis, even post-capitalist solutions. Just as the transition to industrial capitalism, and by implication a properly capitalist society was premised on the internalisation of production – in the shape of the plantation system first and the factory system later – so only the internalisation of reproduction can provide the space for an expanded circuit of accumulation today (not to mention the ecological requirement to account for the hidden costs of capital accumulation by internalising externalities).