The Neofeudalising Tendency of Communicative Capitalism

Jodi Dean

Hobart and William Smith Colleges, Geneva, NY, USA,

Abstract: Communicative capitalism is becoming neofeudal. From network effects, to plaformisation, and the rise of a service-based economy, the tendencies of networked communications are neofeudalising.

Keywords: neofeudalism, communicative capitalism, platforms, tech lords, serfs, hierarchy, symbolic efficiency

1.   Introduction

In 2010, in the influential and prescient book You Are Not a Gadget, Jaron Lanier (who identifies as the father of virtual reality) discussed newly emergent cloud computing in terms of lords and peasants (Lanier 2010). The lords own and control platforms and data. The peasants or serfs are the rest of us who have become dependent on these platforms. Someone else owns the tools we need to do our work; someone else—the platform lord—is the conduit or means through which we access the market; someone else stores our data, charges us a fee for all this, and collects metadata about our transactions and use. Of course, Lanier wasn’t yet able to trace out all these details of platformisation, but he identified the basic structure and dynamics right from the start. Under capitalism the so-called sharing economy was always going to be about someone making money.

Over the next few years, the link between networked information technology and social property relations reminiscent of feudalism solidified. Writing in the Harvard Business Review, Bruce Schneider – a network security expert – concluded a list of nefarious dealings by Facebook, Google, Apple, Microsoft, Twitter, and LinkedIn by pointing to the shift of power to IT. IT’s dramatic increase in power, he said, was indicative of a “digital feudalism“. He warned, “If you’ve started to think of yourself as a hapless peasant in a Game of Thrones power struggle, you’re more right than you realise. These are not traditional companies, and we are not traditional customers. These are feudal lords, and we are their vassals, peasants, and serfs” (Schneider 2013). More recently, Yanis Varoufakis, economist and former Greek finance minister, has added his voice to the growing alarm, declaring “techno feudalism is taking over” (Varoufakis 2021).

Although rejected by critics such as Evgeny Morozov, who initially embraced the techno feudal hypothesis, the idea that we are dependent vassals under the dominion of big tech hardly seems ground-breaking (Morozov 2022). Tech billionaires increased their wealth by over $360 billion in the first year of the pandemic, a time when millions of workers lost their jobs or were forced to stay at their jobs and risk their health and lives. The extreme inequality associated with our tech overlords is well-known. What’s surprising is how different this reality, this timeline, is from the view of networked digital communications twenty-five to thirty-years ago.

2.   Back to the Future

In the early 1990s and even into the early 2000s, the Internet was associated with democracy. Citizen journalism, town halls for millions, transparency in government, and the opportunity for everyday people to make their voices heard was going to usher in a golden age of political participation. Anyone with a laptop or cell phone would be able to get their message out without having to go through the censors, judgments, and hierarchies of gate-keeping institutions. Information would be free. Entrenched power structures would crumble. The chains of tyranny would be lost. A primary buzzword for this democratic fantasy was “participatory“. Attaching “participatory” as a prefix heralded the revolutionary change in social relations the Internet was bringing about: participatory media, participatory budgeting, participatory art, and so on.

In the nineties, the Internet was also associated with friction-free capitalism and a smooth world of flows. Observers as far apart as the neoliberal New York Times columnist Thomas Friedman and the communists Michael Hardt and Antonio Negri shared the same technological imaginary of global horizontal productivity. Networked communications were to enable a New Economy of abundance rather than scarcity. The productivity and creativity of the multitude would exceed and rupture state control.

An oft-repeated slogan announced: information wants to be free! Did this mean free as in unrestricted or free as in no one has to pay for it? If the latter, how were content creators going to eat? The blurring or instability around everything being free on the Internet was indicative of the larger, more fundamental equating of democracy and capitalism that the defeat of the USSR seemed to cement. The same infrastructure that let some get rich in the dotcom bubble – and would support the “killer apps” and big data rush in the 2000s – made everybody freer. Even better: networked personalised media was fun! It turned work into play and users into producers (new media scholars liked to throw around terms like playbor and prosumer, that is, consumer as producer, which really meant producing for capital without being compensated or, differently put, that capital had figured out how to extract value from processes of private consumption and use). At any rate, my point is that in the nineties and even after the dot com bubble burst and we moved to the 2000s’ “web 2.0,” communication technologies made capitalism acceptable, exciting, and cool, immunising capitalism from critique by rendering critics into outmoded technophobes.

3.   Hierarchy is a Feature not a Bug

Thirty years ago, the promise of the Internet was more democracy, abundance, and freedom. That fantasy unleashed a set of dynamics that have intensified inequality, undermined the shared understandings necessary for democracy, and enabled the rise of the far right around the world. Instead of an era guided by communicative action in a democratic public sphere (to use terms from Jürgen Habermas), expansions in networked personal communications entrapped us in communicative capitalism, that era of capitalism where communication is central to production, distribution, circulation, and accumulation (Dean 2009).

To get a sense of the depth and breadth of the merger of capitalism and communication consider:


1.  The complex logistics that support a trade system built on the concentration of industrial production in special economic zones.

2.  The automation and informatisation of productive processes that standardise and accelerate production while decreasing the need for human labour power.

3.  The high-speed networks enabling algorithmic trading, hedging, and arbitrage in financial markets.

4.  The injunction to individuate and the corresponding over-burdening of the individual form as ever more choices and responsibilities are downloaded onto individuals via personalised networked communications devices.

5.  The new capacity for capital to monitor and capture the activities through which we reproduce our social lives (big data, Internet of Things).

6.  “Platformisation” or the rise of platforms the use of which generates data and income for platform owners (Google, Facebook, Amazon, Uber, AirBnB, etc.).

7.  The intensification and spread of surveillance.

8.  The intensification of work and expansion in time spent working, as people are expected to be connected and responsive 24/7. 

9.  The build-up of fixed capital to accommodate data storage and the training of the algorithms that can process and learn from this data as well as generate new algorithms, in other words, the monopoly power of big tech (Rikap 2023).


These features of the contemporary economy rely on the merger of information and communication systems with systems of extraction, production, and distribution. The tech giants – Facebook (Meta), Google (Alphabet), Apple, Microsoft, and Amazon – grow ever richer and more predatory, their founders becoming billionaires because of the cheap labour of their workers, the outsourcing of work to third party contractors, the free labour of their users, the tax breaks bestowed on them by cities desperate to attract jobs, and the solidification of their monopolies. The extractive dimension of networked technologies is pervasive, intrusive, and unavoidable. The tech giants have become masters of the entire process of social life. It might be an exaggeration to say that the present is literally an era of peasants and lords (although Elon Musk talks about Twitter users this way). But it is right to say that contemporary capitalist society is characterised by an intensification of inequality – more billionaires, greater distance between rich and poor, and the solidification of a differentiated legal architecture that protects corporations and the rich while it immiserates and incarcerates the working and lower class. 

The promise we were sold back in the nineties was that networks would lead to a horizontal society, a flat world of interconnections and equal possibility. But as Albert-Láslzó Barabási’s research on complex networks demonstrates, free choice, growth, and preferential attachment – the characteristics that define complex networks – produce extremes of inequality (Barabási 2003). In complex networks, people voluntarily make links or choices. The number of links per item or site grows over time, and people like things because others like them (for example, seeing that some have liked something on Facebook leads others to like it as well). Link distribution in complex networks follows a powerlaw where the most popular item generally has twice as many hits or links as the second most popular, which has twice as many as the third most popular and so on, down to the insignificant differences among those in the long tail of the distribution curve. The one at the top has significantly more than the ones at the bottom. This winner-takes-all or winner-takes-most effect is the powerlaw shape of the distribution. The shape the distribution takes is not a bell curve; it’s a long tail – a few billionaires, a billion precarious workers.

Complex networks encourage inclusion. The more items in the network, the larger the rewards for those at the top. Growth is a characteristic as well as an aim. Complex networks also induce competition – for attention, resources, money, jobs – anything that is given a network form. And they lead to concentration: hubs, monopolies, blockbusters, influencers. The result, then, of free choice, growth, and preferential attachment is hierarchy, powerlaw distributions where those at the top have vastly more than those at the bottom. What this means is that complex networks have neofeudalising tendencies. Inequality and hierarchy – powerlaw distributions – are built in, features not bugs. Their basic logic produces lords and peasants, one and many.

Although often overlooked by tech writers, the neofeudalising tendencies of communicative capitalism show up in the servant economy. I’m not referring here to the replacement of jobs by automation but rather its opposite: the limits of automation. Jason E. Smith draws out Marx’s analysis of the connection between industrialisation and the expansion of services (Smith 2020). As productivity increases, requiring fewer workers, those in need of a wage to survive are thrown into sectors less amenable to automation, that is, services. Services are less amenable to automation in part because of the specific skills care work requires, diapering a baby, moving an elderly person from bed. They also resist automation because they are cheap, the last jobs available to those thrown out of every other. As Marx writes in Capital, Volume I: “the extraordinary increase in the productivity of large-scale industry, accompanied as it is by a more intensive and a more extensive exploitation of labour-power in all other spheres of production, permits a larger and larger part of the working class to be employed unproductively. Hence it is possible to reproduce the ancient domestic slaves, on a constantly expanding scale, under the name of a servant class” (Marx 1867/1976, 574). Capitalist industry itself reproduces social property relations characteristic of earlier economic forms.

Society is no longer oriented toward the production of workers and commodities. Its’s now an order of personalised service, privilege, hierarchy, and fealty – not a social factory but a social manor. More and more of the people forced to sell their power to survive sell this as services to those looking for deliveries, drivers, cleaners, trainers, home health aides, nannies, guards, coaches, and so on. The buying and selling of services is enabled by new intermediaries, technological platforms whose owners insert themselves between service offerors and seekers, being sure to exact a fee along with the data and metadata that accompany the transaction. Our basic interactions are not our own. With advances in production seemingly at a dead end, capital today is hoarded and wielded as a weapon of disruption, its wielders new lords, the rest of us dependent, proletarianised servants and serfs.

4.   The Social Manor

Although there has been widespread land-hoarding on the part of the super-rich, the primary mode of surplus extraction today is not a landed aristocracy sitting on a peasant class legally bound to the land (it should nevertheless be noted that the largest landowner in the world is the King of England; the British royal family owns more 6.6 billion acres of land). Neofeudalism does not designate the simple return of the feudal form of exploitation (Dean 2020). In the so-called advanced economies, the majority of people do not have direct access to their means of reproduction. They require some kind of income that will enable them to purchase what they need to subsist – food, shelter, and means of life. Neofeudal serfs are proletarianised serfs, “free” from the land, “free” from job security, “free” from social welfare safety nets, and dependent on markets for every aspect of their lives.

Networked telecommunications have enabled the insertion of a new appropriating class at precisely this point – access to markets. This is what platforms like Google, Facebook, Amazon, Apple’s App Store, Uber, and Airbnb provide – “digital infrastructures that enable two or more groups to interact” (Srnicek 2016, 43). Platforms position themselves between buyers and sellers, seekers and suppliers, mediating interactions while collecting fees and data. Even more, their monopoly power and enormous infrastructures produce ecosystems that shape economic life in ways impossible to escape.

Even though technology companies employ a relatively small percentage of the workforce, their effects are tremendous, essentially remaking entire industries around the acquisition, mining, and deployment of data. In fact, it’s because of the smaller workforces that data tech is indicative of a tendency to neofeudalism. Capital accumulation occurs less through commodity production and wage than through services, rents, licenses, fees, work done for free, and data treated as a natural resource. Those who work for a wage are subjected to ever-worsening conditions as algorithmic scheduling and surveillance tech press them to work harder and faster, with few breaks and fewer benefits. For workers in Amazon warehouses, call centres, and fast food, work is degrading, soul-sucking, abusive, and literally dehumanising: that people tire, use the toilet, can’t always work to full capacity, and have lives isn’t part of the equation (Guendelsberger 2019).

Crucial to the power of platforms is the way they constitute grounds for user activities; platforms are the conditions of possibility for interactions to occur. Google makes it possible to find information in an impossibly dense and changing information environment. Amazon lets us easily locate, compare prices, and purchase consumer goods from established as well as unknown vendors. Uber enables strangers to share rides. Airbnb does the same for houses and apartments. One’s car isn’t for personal transport. It’s for making money. One’s apartment isn’t a place to live; it’s something to rent out. Personal property becomes an instrument for the capital and data accumulation of the lords of platform. The more people use these platforms, the less we can avoid. They become more effective, more powerful, ultimately transforming the larger environment of which they are a part.

Instead of rearranging production processes and implementing technological improvements to increase productivity, what Marx referred to as real subsumption, capital accumulation increasingly takes place under conditions of formal subsumption. Non-capitalist processes produce value that capital, whether via measures of finance or governance, simply takes. Negri (2019) associates the resurgence (or non-weakening) of full subsumption with producers’ reappropriation of fixed capital in ways that let them autonomously experiment with organising production in common. More compelling examples of resurgent full subsumption are Uber and Airbnb. Neither create new efficiencies in the provision of local transportation or lodging for travellers. Both rely on the fixed capital of earners rather than on capitalists who invest in means of production. Both turn assets acquired out of workers’ consumption funds – their cars and houses – into means for the capital accumulation of another, the corporation.

Amazon further illustrates the primacy of monopoly predation and control indicative of communicative capitalism’s neofeudalising tendency. Stacy Mitchell describes Amazon as a toll road (Mitchell 2021). It extracts enormous fees from the merchants wanting to sell goods in its marketplace. These include the fee to sell on the site, called a referral fee, and two “optional fees” – advertising and fulfilment. Because of Amazon’s practice of prioritising sponsored ads in search results, sellers feel compelled to buy additional advertising (instead of relying on customer ratings to elevate them in search results). Furthermore, many sellers have turned to Amazon for warehousing and shipping, Fulfillment by Amazon (FBA). Again, this is because the algorithms prioritise FBA sellers in search results. Selling on other sites is barely an option: not only does Amazon’s domination of the online sales market render other sales outlets less viable and harder to find, but sellers can’t lower their prices on these other outlets. Amazon’s “competitive pricing” rules demote the items in search results, eliminate the “buy now” button, and add an alert that tells customers that the item is priced higher on Amazon than elsewhere.

Amazon’s aggressive fee structure enables additional predatory practices. Amazon loses billions with the free shipping promised by its Prime subscription service. The popular service, though, is what lets Amazon maintain its dominance in online sales. Customers have strong incentives to shop on Amazon when they’ve paid for Prime. The enormous sums that seller fees generate more than makes up for the Prime losses. In fact, the fees for advertising, warehousing, and shipping paid for Amazon’s enormous logistics build up. In effect, Amazon pushed the costs onto its sellers. They built the infrastructure they had no choice but to use.

Amazon’s fees and tolls combine with the rent-driven accumulation strategy of Amazon Web Services (AWS). Together with Azure (Microsoft) and GCP (Google), AWS is a cloud platform. Users are charged for processing (called “compute”), networking (input and output), and storage. Instead of building up their own technological infrastructures, firms purchase computing services from larger tech companies. The big three tech companies control about sixty-five per cent of the cloud market. If a smaller competitor poses a risk, all they have to do is buy it. Functioning like public utilities in that they provide essential services too expensive for most firms to supply for themselves, the structural dominance of these companies enables them to act in predatory and unaccountable ways. With the development of machine learning and generative AI (artificial intelligence), the power asymmetries seem insurmountably immense. Developing and training deep learning applications requires enormous data sets and amounts of compute time. Only a few companies are in a position to provide much less afford it. OpenAI – which developed and released the ground-breaking AI ChatGPT – uses Azure; Microsoft invested ten billion dollars in the company, acquiring a forty-nine per cent share.

Lordship designates a social relation for the appropriation of surplus. Are we not in that same relation with the big tech? The theme of peasants and lords of the Internet earlier noted by Lanier has escalated to become widely accepted common sense as the tech giants have grown richer and more extractive because their structural dominance enables them to extract increasingly unavoidable tolls, fees, and rents. Neither consumers nor businesses can avoid them. Cities and states compete to attract them. The economic scale and impact of our tech overlords is greater than that of most so-called sovereign states. Like so many tributary demands, their tax breaks take money from communities. Their presence drives up rents and real estate prices, driving out affordable apartments, small businesses, and people. Shoshana Zuboff’s study of “surveillance capitalism” brings out a further neofeudal element of tech lordship – military service (Zuboff 2019). Like powerful lords to needy kings, Facebook (now called “Meta Platforms”) and Google (Alphabet) cooperate with state governments, sharing information that these states are legally barred from gathering themselves.

5.   A Tendency Not a Law

Neofeudalism is not inevitable. I am not arguing that it is impossible for the Internet to take any other shape or that there are no other ways to structure networks. Social networks can be designed to counter neofeudalising tendencies. To counter preferential attachment, for example, sites could not report how many people like something. They could not recommend things to us because others like them. They don’t have to show how many followers someone has. Social networks choose how to rank or privilege posts. It’s a matter of algorithm design. Before 2009, the Facebook feed was chronological. Then the algorithms were adjusted around popularity, and after that they were individualised – these are choices (Teachout 2022). Tik-Tok takes preferential attachment to an extreme, combining several preference measures – time watching, likes, and comments – to determine which videos to push to whom and create powerlaw distributions, videos that stand out from the rest and go viral (Smith 2021).

Likewise, the commercialisation of the Internet was not inevitable. The US could have worked through the UN or with other countries to develop a fully public Internet. It didn’t have to transfer responsibility for supervising the Internet backbone from the National Science Foundation’s NSFNet to private corporations and it certainly didn’t need to deregulate telecommunications in 1996. That deregulation paved the way for the mergers and consolidations that concentrated media ownership into massive conglomerates. Bluntly put, networked communications didn’t have to take the form of private property. It could have been protected as a commons or public utility. Instead of the dominant search engine being the creation and property of a corporation, it could have been publicly developed and funded, like the Internet was initially.

This leads to one last example of something that was not inevitable: the collection of data and metadata and its treatment as a natural resource available for private appropriation. Corporations could have been prevented from collecting this data, from storing it, from selling it. There could be limits on server sizes and server farms (which would be environmentally significant given how much energy they use) to disincentivise data collection. All these – algorithm design, commercialisation, data gathering and storage – could have been and could still be different, but that’s a political choice. It requires political will to change them, political will to undertake an enormous fight.

So, where’s the fight? If networked communications have led to the intensification of monopoly power and widespread inequality, where is the resistance? To be sure, people criticise networked media. Zuckerberg and other tech lords have had to testify before Congressional committees. People love to hate the tech billionaires and point out the dangers of Instagram and TikTok. Liberals get especially incensed when X (Twitter) and Facebook allow the uncensored circulation of misinformation. But at best liberals call for regulation. They don’t demand that the tech giants be seized and turned into public utilities. And the basic reason why is that mainstream politicians accept capitalism. They don’t think that in principle it’s wrong for there to be billionaires and media conglomerates with the power to scrape up, own, mine, and sell our personal information – and to charge us for doing so.

If the neofeudalising effects of communicative capitalism are so bad, why aren’t millions of people in the streets demanding change? Why hasn’t a significant movement for socialising the Internet emerged, a movement of tens of millions rather than a few thousand? Where is the revolutionary consciousness? The answer has to do with the way that networked personalised digital media impacts communication and hence consciousness. Most broadly put, our communication networks are primarily affective networks, networks oriented toward the circulation of outrage. They fragment understanding and undermine meaning, contributing to a world steered by images that resist consolidation in explanations and organisations with a capacity to endure.

Consider three changes in communication characteristic of communicative capitalism. The first is the shift from message to contribution, a shift from the use value to the exchange value of an utterance. A basic way to formulate the idea of communication is in terms of message and response. A message is delivered to a receiver with the intention of eliciting a response. This changes in communicative capitalism. Messages are contributions to circulating content, not actions to elicit responses. The exchange or circulation value of the contribution overtakes the use value of the message. Meaning, or use value, matters less than the likelihood that a contribution will be shared. Whether a contribution is a lie doesn’t matter. What matters is simply that something was expressed, that a comment was made, that an image was liked and shared. Communication is subjected to an economic logic.

An example is being “ratioed” on X (Twitter) – this refers to the presence of a high number of comments on a tweet. Typically, a high number of comments means a high level of disagreement (in contrast to retweets which generally, but not always, indicate agreement). The number, the aggregate, is what registers; the actual content of any given comment, or even the fact that the various comments say widely different things, is eclipsed. And of course, a single tweet’s ratio itself gets eclipsed by what trends, which is usually forgotten in a matter of hours. In social media, something well-argued, true, and important to a matter of real concern rarely or barely registers because the stream of contributions is endless, constant. Something else that is true and important will not just appear tomorrow but is appearing at the same time, in the same feed, making the same demands for attention. As equal contributions to circuits of information and affect, then, the content of our utterances is unimportant. What matters is their mobility, their capacity to circulate. Dissent is just more content, whether cogently argued or the daily outrage. What’s the point of critique if it won’t, if it can’t, register? Communicative capitalism undermines critique’s conditions of possibility.

The second change involves the affective character of networked communication. The channels through which we communicate reward number: the more hits and shares the better. It’s now clear to everyone that the circulation capacity of outrage vastly exceeds that of cogent arguments. Reaction time is shortened; responding is easier. In a setting of constant, unceasing, infinite, and ever-intensifying demands on our attention, we don’t have time to respond to everything, to evaluate everything, to reflect. Nuance takes too long. Science doesn’t register but intensity does, whether as outrage, absurdity, cuteness, or reassuring clichés (which are easy to recognise, making at least some people feel at home, like, finally, they, too, get what social media is about). You don’t argue with a meme; you rant through a meme, hoping to impact someone enough that they will share it or copy it. But even this will be transient, a momentary dopamine hit that won’t last long and that will dissolve in the larger flow.

The networks of communicative capitalism are affective because images and emotions circulate more rapidly than ideas. It’s easier to share the photo of the scared koala on the edge of a burning forest than it is to read, digest, summarise, and share an article about Australia’s increased commitment to fossil fuel extraction. It’s easier because one’s decision about sharing can happen rapidly, and one can expect that others will affirm this decision with likes and shares. The repercussions of the communicative practices of affective networks are not confined to digital networks.

Why did people ever think that a communication terrain promising to include everyone and enable anyone to say anything at any time would be a good idea? Everyone includes hucksters and trolls as much as it does female-identified artisans and sincere rural teenagers. It includes flat-earthers and fascists as much as it does quantum physicists and communists, supporters of apartheid as well as supporters of liberation. The material conditions of analogue media reduce and slow down encounters between wildly divergent worldviews; digital media bombards users with them. When it took effort to tune into the controversial radio show or locate the banned book, elites and majorities could ignore them and subcultures could embrace them. Today strange and unsettling takes sweep like wildfires through our communicative habitats, destroying the assumptions and practices necessary for meaning in the process.

In digital communication networks, we regularly confront myriad others whose views of reality differ from our own. Our disagreements are not just matters of taste and opinion. They’re not even just about morality and the good life. Disagreements are about reality itself. Again, these have always been there, but networked digital communication brings us into speeded up contact with them: denialists of all stripes, influencers whose every upload tries to sell us something, an endless array of cynics and nihilists. All news is fake to somebody – and we find ourselves pressed to react to all of it, right here, right now.

Third, digital communication has resulted in a decline of symbolic efficiency (an expression I get from Slavoj Žižek). Symbols, points of reference, that signify one way in one field mean something altogether different in another one – and, there is no way to stabilise meaning or provide a decisive determination (the big Other doesn’t exist; division goes all the way down). In the US, the pandemic and the 2020 election forced us to confront this decline in symbolic efficiency. There is no big Other who can tell us whether the virus is real, masks work, vaccines are safe, and who won the election. These are not merely political disagreements; the political disagreements arise from the material change in communication. The collapse of the symbolic is suffocating, a kind of closure that makes significant action impossible because it can’t register. In fact, the politics itself isn’t clear; the pandemic and the 2020 election blur left and right– our setting is not one of symbolic efficiency where we can map the sides with any certainty.

Some users have adapted to the disorienting cacophony of the decline of symbolic efficiency through careful curation of their feeds. Already in his 1995 book Being Digital, Nicholas Negroponte predicted that networked media would enable users to pick and choose what sort of information they want to consume. He worried that this could diminish encounters with unexpected, unsearched for, stories. Thirty years later – after enduring flame wars, trolls, Gamergate, bullying, gas-lighting, streams of lies and hate, not to mention privacy violations and countless ads – social media users tailor their lists of friends and followers, hoping for interactions that won’t amplify the low-level outrage that’s become communicative capitalism’s primary affect. Even as some call out for Facebook and X (Twitter) to do this culling for them, to protect them from fake news as if these capitalist corporations should determine for us all the meanings of “fake” and “news”, many form groups, tribes, and bubbles, loci of assurance. They seek out those whose interests or experiences may be similar, who’ve faced the same kind of challenges, be they challenges around a complicated recipe, navigating bureaucracies, or dealing with trauma. The necessity of these affiliations is attested to by the vehemence with which they are defended: from competing fandoms to the reflex to cancel, defending the fortress that protects who one is has become an imperative, experienced as a matter of life and death.

The politically engaged also seek out allies, comrades, and fellow travellers. Some liberal analysts criticise these “information silos“. Operating as if the ideal of a public sphere were operative online, they encourage users to seek out those with whom they disagree, engage them, and look for common ground. One wonders if these liberals are disingenuous or have never been online: networked digital interactions with those outside one’s bubble lack the shared understanding necessary for political discussion. Reality is not the same. Words mean different things. Siloing, bubbles, make sense as a reaction to the absence of shared meaning in digital networks. Like-minded comrades provide necessary support against the ceaseless waves of outrage. Fragmentation, polarisation, is a mass adaptive response to the decline of symbolic efficiency, not an individual failing.

In the affective infrastructures of communicative capitalism there is no commonly shared meaning. With constant communication but maddening incommunicability, we lose the sense not just that no one hears us, but that even if they do there is no way they can understand us. Meaning is individuated; what matters to me, what I have experienced, what I feel. Under these conditions of the decline of symbolic efficiency, disparate issues and concerns are equalised – the daily deaths of thousands circulating in the same media space as fury over a language faux pas and straight up lies. Little registers as significant – or everything is equally significant – because signification has become so elusive. Cries for help don’t register; demands for change seem utterly inadequate – especially because someone will immediately contradict or deflect them. We get overloaded by pervasive feelings of insecurity, anxiety, and catastrophe, overloaded because there’s no relief, no redress, no escape.

The affective networks of communicative capitalism provide a habitat more conducive to neofeudalism than emancipatory egalitarian revolution. Fragmentation as an adaptive response to the decline of symbolic efficiency helps explain some of the challenges involved with building class consciousness today. Not only have the conditions of characteristic of the height of industrial unionisation changed due to offshoring, automation, and the rise of services, but the shared reference points and cultural commonalities necessary for communication have ruptured. In the most advanced economies in the world, the conditions of the majority of workers resemble the isolation and idiocy Marx associated rural peasants, idiocy born not from stupidity but from a narrow confinement to the particular broken only by incessantly circulating outrage.

In this setting, the primary language of politics has devolved into safety and survival. Anything more seems utopian, unrealistic on a warming planet and impossible to organise even if we could imagine it. Describing the average US American as a powerless “neo-serf”, Umair Haque sees “a creation unique in modern history – a nominal citizen of a rich country, but one whose every imaginable possibility goes on shrinking by the day. [...] because he is impotent, castrated, every day can only feel something like either torment or death” (Haque 2022). People have good reasons for feeling insecure. The catastrophe of neofeudal capitalism’s expropriation of the social surplus in the setting of a grossly unequal and warming planet is real.

6.   Conclusion

Communicative capitalism is tending to neofeudalism. The shift from message to contribution, affective character of communicative networks, and decline of symbolic efficiency impact the kind of people we are, consciousness we have, and communities we can create. Communication is a material dimension of the production and reproduction of life. Now more than ever it’s part of the base as well as the superstructure. It’s the infrastructure connecting, bridging, meshing base and superstructure. This grim story reminds us that the hierarchical character of complex networks supports the private accumulation of the few, not the flourishing of the many. It tells us that the problems of fake news and anti-vaxxers and all that is pushed on us as new and different are features baked into digitally networked communications. It attunes us to the limits of critique in a setting where truths and lies are communicatively equivalent. And it should show us why our politics today must be militant and focused on building tight, disciplined, and committed political organisations.


Barabási, Albert-Láslzó. 2003. Linked. New York: Plume.

Dean, Jodi. 2020. Communism or Neo-feudalism? New Political Science 40 (1): 1-17.

Dean, Jodi. 2009. Democracy and Other Neoliberal Fantasies. Durham, NC: Duke University Press.

Guendelsberger, Emily. 2019. On the Clock. New York: Knopf.

Haque, Umair. 2022. “Why the World is Going Backwards – And How to Stop It“. Medium.

Lanier, Jaron. 2010. You Are Not a Gadget. New York: Knopf.

Marx, Karl. 1867. Capital. Volume I. London: Penguin.

Mitchell, Stacy. 2021. Amazon’s Toll Road. Institute for Local Self-Reliance.

Morozov, Evgeny. 2022. Critique of Techno-feudal Reason. New Left Review 133/134: 89-126.

Negri, Antonio. 2019. The Appropriation of Fixed Capital: A Metaphor? In Digital Objects, Digital Subjects: Interdisciplinary Perspectives on Capitalism, Labour and Politics in the Age of Big Data, edited by David Chandler and Christian Fuchs, 205-214. London: University of Westminster Press.

Negroponte, Nicholas. 1995. Being Digital. New York: Knopf.

Rikap, Cecilia. 2023. Capitalism As Usual? Implications of Digital Intellectual Monopolies. New Left Review 139: 145-160.

Schneider, Bruce Schneider. 2013. You Have No Control Over Security on the Feudal Internet. Harvard Business Review, June 6.

Smith, Ben. 2021. How TikTok Reads Your Mind. The New York Times, December 5.

Smith, Jason E. 2020. Smart Machines: Automation in an Age of Stagnation. London: ReaktionBooks.

Srnicek, Nick. 2016. Platform Capitalism. Cambridge: Polity Press.

Teachout, Zephyr. 2022. When Big Brother is Your Boss: The Rise of Surveillance Wages. The Nation, January 5.

Varoufakis, Yanis. 2021. Techno-Feudalism Is Taking Over. Project Syndicate, June 28.

Zhang, Mary. 2023. ChatGPT and OpenAI’s use of Azure’s Cloud Infrastructure. Dgtl Infra, January 26.

Zuboff, Shoshana. 2019. The Age of Surveillance Capitalism. New York: Public Affairs.

About the Author

Jodi Dean

Jodi Dean is Professor of Politics at Hobart and William Smith Colleges in Geneva, NY. She is the author of numerous books including Democracy and Other Neoliberal Fantasies, Durham, NC: Duke University Press, 2009; The Communist Horizon, London: Verso, 2012; Crowds and Party, London: Verso, 2016; and Comrade: An Essay on Political Belonging, London: Verso, 2019. She is currently finishing a book on neofeudalism.