SciELO - Scientific Electronic Library Online

vol.54 special issue author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand



Related links

  • On index processCited by Google
  • On index processSimilars in Google


Acta Academica

On-line version ISSN 2415-0479
Print version ISSN 0587-2405

Acta acad. (Bloemfontein, Online) vol.54 spe Bloemfontein  2022 



Algopopulism and recursive conduct: grappling with fascism and the new populisms vis-à-vis Arendt, Deleuze and Guattari, and Stiegler



Chantelle Gray

Department of Philosophy, Northwest University, Potchefstroom. ORCID:; E-mail:




The continuing rise of right-wing populisms and fascism across the globe, mobilised by various factors such as the use of social technologies and the weaponisation of nationalism, xenophobia, sexism and racism - to name a few - provokes important questions in terms of resistance. Moreover, given that the lines between the new populisms and fascism are becoming increasingly nebulous, it is imperative that we at least attempt to better understand the conditions - including the more hauntological, which is to say historically and structurally invisibilised - from which algorithmic governmentality and its corollary, recursive conduct, are born. On that account, I unpack two understandings of fascism, first tracing it macropolitically through the work of Hannah Arendt and, thereafter, looking at its more micropolitical and libidinal aspects vis-à-vis the work of Deleuze and Guattari. I do so because it is my contention that there is a historical correlation between Arendt's theorisation of a generalised espionage and our contemporary surveillance systems, as well as between Deleuze and Guattari's theorisation of desire and the harnessing of affect through machine learning methods and their deployment via social media platforms, all of which aid the propagation of certain forms of populism - and even fascism. This is what I call algopopulism: algorithmically aided politics that transforms the 'we' into the 'they' through what can be thought of as recursive conduct or the digital exercise of power and its structuring of the field of possible action and thought. In turn, I link this to Stiegler's understanding of negative sublimation, a paralysis of the human spirit which occurs due to, among other things, the generalised proletarianisation of knowledge, ultimately provoking the short-circuiting of processes of transindividuation. Finally, I offer some notes on resistance.

Keywords: fascism, algopopulism, micropolitics, recursive conduct, algorithmic governmentality



Three vignettes (an oblique introduction)


Many readers may recall the first episode of the third season of Black Mirror, titled "Nosedive" - the one in which social scorecards have been embraced and normalised in society. In the opening scene, viewers watch as Lacie Pound (Bryce Dallas Howard) gives someone's post a five-star rating, quite like the way you might rate an Uber driver. Smiling, she walks towards a barista to collect a coffee. His rating pops up: 3.7. Average. Lacie, on the other hand, is a 4.2. Not bad, but she needs to up her game if she is to qualify for a discount on a luxury apartment. This, viewers soon find out, means adjusting her behaviour just ever so slightly: adding a few fake smiles, giving a couple of phony compliments, maybe even repressing feelings of anger and depression at times. Not all the time, mind you, but there is a certain amount of people-pleasing involved which, frankly, is not such a bad thing, right? We could all do with a bit adjustment here and there - especially if it pays off in certain tangible ways like qualifying for discounts. Right?! But what happens when you lose your temper, when things snowball on a day and you just can't seem to find your zen? What about the inevitable nosedive?

Such social credit systems, as we know, are not distant sci-fi inventions. We have many 'soft' versions of them all around us, for example liking Facebook posts and rating Airbnb hosts. 'Harder' versions also exist, for example the social credit system in China - "a broad policy project for encouraging individuals, businesses, legal institutions, and government itself to be more 'trustworthy' (守信, shouxin) through a mix of measures" (Ahmed 2019; emphasis added). Like the social credit system in Black Mirror, the one in China is not quite as nefarious as the many gross mischaracterisations thereof would have us believe. Rather than being a "digital dictatorship" where "big data meets Big Brother", it is a "cluster of policy initiatives" intended to encourage and incentivise improved behaviour to "promote greater trust" (Ahmed 2019). That said, social credit systems - which can include anything from the creation of black lists (for corrupt behaviour) to red lists (for upright behaviour) to 'participatory' refugee confinement systems, often used by well-intentioned NGOs to gather information/data on refugees, ostensibly for programmes to aid their wellbeing - do often pre-establish and surreptitiously impose the "goals and forms of participation" in advance (Tazzioli and Pârvan 2021: 1), which raises at least some questions about the voluntariness of activities, the relation between knowledge and opacity, surveillance, and interpellation more broadly. In other words, whether gross mischaracterisation or not, social credit systems are mechanisms of discipline and control. But are they fascist?


The meme read "Steal his look". On the left-hand side was a picture of the now infamous shirtless man known as the 'QAnon Shaman', so named for the attire he wore when storming the US Capitol. The rest of the meme had pictures of his 'shamanic' clothing arranged as items you can buy: super-soft fantasy horns ($11,95), coyote fur Russian hat ($279.95), duck dungaree ($34.95), eyeshadow pallet ($29.00), pendant necklace ($60.00), 20-watt megaphone ($14.99) and touchscreen gloves ($54.99). Priceless.

Variously known as Jake Angeli and Yellowstone Wolf, Q Shaman - donning a horned fur hat, wearing 'war' paint in the colours of the US flag, and carrying a spear with the American flag tied to it - participated in the US Capitol attack on 6 January 2021 in support of ex-president Donald Trump who had piteously been spreading lies that the 2020 election had been stolen by the 'radical-left' Democrats. (The left, meanwhile, were wondering how the Overton window had shifted so far to the right that Joe Biden was considered left at all.) It was a bleak January, to be sure. Not just because Trumpism was still in the air, but also because the world had just entered the second year of the Covid-19 pandemic.

Of interest about Q Shaman is less his habiliments (shocking as this may seem) than his being an adherent of QAnon, "a decentralized, ideologically motivated and violent extremist movement rooted in an unfounded conspiracy theory that a global 'Deep State' cabal of satanic pedophile elites is responsible for all the evil in the world" (Argentino 2021). Like most conspiracy theories, QAnon draws on some real-world cultural and sociopolitical uncertainties, but distorts and simplifies these by integrating them into an ideological schema. It is not surprising, then, that the Covid-19 pandemic, so ridden with uncertainty, "played a significant role in popularising the QAnon movement. Facebook data since the start of 2020 shows QAnon membership grew by 581 per cent - most of which occurred after the United States closed its borders" as "part of its coronavirus containment strategy" (Argentino 2021). Despite Trump's hand in the mismanagement of the pandemic in the US, he is cast in a messianic role in the conspiracy theory because it was Trump who "finally gave QAnon what it always wanted: respect" (Argentino 2021). This is why, despite the absurdity of the storming of the US Capitol in the name of Trump, it also makes a strange kind of sense - because it was nothing less than the "culmination of what" had "been building up for weeks: the 'hopeium' in QAnon circles that some miracle via Vice-President Mike Pence and other constitutional witchcraft would overturn the election results" (Argentino 2021). But is it fascism?


As the world begins to try to 'live with' Covid and 'normal' life resumes, albeit in somewhat altered ways, many people are trying to come to terms with the past couple of years, to contextualise where we are and where we might be headed. But with so many opposing views it is harder to do than one may imagine. On the one hand, there are people like philosopher Giorgio Agamben who, in the early months of the pandemic, wrote a post in which he claimed the Covid-19 pandemic to be "an alleged epidemic of coronavirus" met with "frenetic, irrational and entirely unfounded emergency measures" - little more than an excuse for governments around the world to "use a state of exception as a normal paradigm for government" (Agamben 2020a). Given what we know about 9/11 and the subsequent 'war on terror', which was used to rationalise the massive centralisation and expansion of state power and securitisation, this kind of claim does not seem entirely unfounded, albeit premature. Agamben would go on to write several more posts of this kind, declaring vociferously in one that the vaccine is "a sort of political-religious symbol aimed at creating discrimination among citizens" through "fascist jargon" that likens vaccine-hesitant people to "bearers of a virtual yellow star" (Agamben 2021). Although many philosophers reacted strongly against Agamben, notably Jean-Luc Nancy, even those who were sympathetic at first drew the line here because, as Benjamin Bratton pointed out, each post became "more absurd and strident than the last" (Bratton 2021). Bratton, in fact, compared Agamben's last-mentioned post to that of Marjorie Taylor Greene, a QAnon-influenced US congressperson who had tweeted something very close to Agamben's sentiments a few months prior to him. Had Agamben surreptitiously fallen prey to right-wing populist sentiments? Had he unwittingly been "unable to account for the epidemiological reality of mutual contagion" because he was too tightly bound "to an untenable set of formulas, reflexi vely suspicious of purposeful quantification" and, as a result, incapable of articulating "an ethics of an immunological commons", as Bratton (2021) claims? Might he perhaps have fallen prey to the power of algorithms and the ways in which they prioritise certain kinds of content based on search histories to individualise social media feeds? Or was Covid-19 indeed a 'plandemic' as the documentary by the same name released in the first few months of the pandemic heralds? And is it fascism? And if so, who are the fascists?

There are no simple answers here but, as I will argue, each of these three examples are, at least in some way, entangled with populism and fascism. To understand how, it is useful here to turn to work by philosophers and political theorists who have grappled with fascism and the new populisms.


A generalised mode of espionage: Arendt on totalitarianism and the spy

Accounts of the holocaust continue to haunt notions of the human as the pinnacle of creation, the recent banning of Art Spiegelman's Maus in Tennessee notwithstanding. And so they should. The systemic, systematic and state-funded oppression and murder of six million European Jews over roughly 12 years was an atrocity - like the middle passage and the slave trade - that we should not be wont to forget. Both spectral and temporal, they are hauntings, in Derrida's sense, which denotes not a belief in spectral beings but "an ethical injunction insofar as it occupies the place of the Levinasian Other: a wholly irrecuperable intrusion in our world, which is not comprehensible within our available intellectual frameworks, but whose otherness we are responsible for preserving" (Davis 2005: 373). Hauntology is, then, the work of memory, at least in part, for memory enfolds within it temporality as well as non-contemporaneity, that which is neither here nor not-here, "a trace that marks the present with its absence in advance" (Derrida 2002: 117). To be sure, hauntology is a ghostly affair, a disquieting phantasm, but it is precisely, if perversely, from the discomfort of the Other that ethical supplication begins. A spectrography, likewise, is a ghostly narrative, a way of perceiving that which exceeds what is representable even as the ineffable impresses itself upon us and implores us to act. This makes hauntology an important lens because it troubles the structuring logics of the absence of history in the present, and the presence of history in what is absented. I want to suggest here, accordingly, that we view Hannah Arendt's account of fascism and its relations to antisemitism, imperialism and totalitarianism in The Origins of Totalitarianism (1962) as a spectrography because, in grappling with the unprecedented - yes, the ineffable - she aims to situate fascism historically, but does so hauntologically by showing how the figure of the spy becomes a "structuring structure", to use Bourdieu's words (Bourdieu 1980, cited in McQuillan 2022: 2), thus figuring as a normative apparatus of governmentality. It is true that Arendt does not describe her own project in this way - she at no point lays claim to hauntology - but I want to resituate her work as such to emphasise its spectral importance through a deconstructive logic that provokes a coming back to, "a structural openness or address directed towards the living by the voices of the past or the not yet formulated possibilities of the future" (Davis 2005: 379).

Born into a German-Jewish family, Arendt was forced to leave Germany in 1933 and lived in Paris for the next eight years, working for a number of Jewish refugee organisations. Perhaps because of her direct experience of antisemitism, she became fascinated by the Jewish question which she saw as the "catalytic agent" for the rise of the Nazi movement and the "establishment of death factories" (1962: viii). Yet she did not think existing political and moral frameworks adequate for grappling with this question, nor its relations to totalitarianism and fascism, which she saw as representing a break in history that had rendered meaningless the existing schemas for grappling with politics, morality and ethics. In her work on fascism and totalitarianism, she thus attempts to provide a novel framework by developing "a new set of philosophical categories" that "illuminate the human condition and provide a fresh perspective on the nature of political life" (d'Entrèves 1993: 2). Importantly, fascism should not be conflated with totalitarianism, even though they share many similar characteristics. In fact, it is useful to distinguish between authoritarianism, marked by strong central governments, such as Hugo Chavez's in Venezuela and Fidel Castro's in Cuba; totalitarianism, characterised by extensive government control in both the public and private spheres of a society, for example Russia under Stalin, China under Mao Zedong and North Korea under Kim Jong-un; and fascism, which is an extreme form of government that combines the worst aspects of authoritarianism and totalitarianism and is typified by military power and a secret police force, as in Mussolini's Italy and Hitler's Germany (Longley 2021).

In her account of the origins of totalitarianism and its culmination in fascism, Arendt argues that the rise of antisemitism "was preceded by a series of financial scandals and fraudulent affairs", one of which was the so-called 'Dreyfuss affair' (1962: 36). The scandal began when Alfred Dreyfuss, "a Jewish officer of the French General Staff, was accused and convicted of espionage for Germany" at the end of 1894 and sentenced to Devil's Island for life (Arendt 1962: 89). The importance of the story for Arendt is not just the Jewishness of the officer, nor the fact that the trial took place behind closed doors, nor even that Dreyfuss was later found innocent, though of course all of these aspects are important. What she finds peculiar, rather, is the newly placed importance on the figure of the spy. As she says:

In a system of ubiquitous spying, where everybody may be a police agent and each individual feels himself [sic] under constant surveillance; under circumstances, moreover, where careers are extremely insecure and where the most spectacular ascents and falls have become everyday occurrences, every word becomes equivocal and subject to retrospective 'interpretation' (Arendt 1962: 431).

To state it differently, rather than being relegated to the sphere of the secret police, the figure of the spy becomes an internalised mode of being because the generalised sense of being spied upon renders suspicion a mechanism for survival and spying the assumed duty of every national, whether at home or abroad. The figure of the spy thus constitutes the practical realisation of the totalitarian state because it is spying that enables the "real secret" of the totalitarian state - namely "the concentration camps" and other similar "laboratories in the experiment of total domination" - to be shielded "from the eyes of their own people as well as from all others" (Arendt 1962: 436). This, Arendt proffers, is one of the central factors for her arguing that "totalitarian rule confronts us with a totally different kind of government" (Arendt 1962: 461). The difference, engendered by the figure of the spy, is that totalitarianism mobilises isolation through a generalised mode of espionage which dispossesses people of both solitude and companionship, rendering action and mutual aid impossible. In her words: "Isolation may be the beginning of terror; it certainly is its most fertile ground; it always is its result" because "the iron band of total terror leaves no space for" a "private life" just as surely as "the self-coercion of totalitarian logic destroys" a person's "capacity for experience", thought and action (Arendt 1962: 474). That is, for a fascist regime to function, it has to transform classes into masses which is predicated on isolation because classes still allow for human connection and a sense of comradery by way of class alliances. It is not enough, therefore, to shift the centre of power from the military to the secret police to realise fascism. Rather, the entire nation needs to function as the secret police because that alone is what ensures mass isolation and, concomitantly, mass obedience. It is the transformation in subjectivity to one of a generalised espionage that is the genius of Arendt's analysis, and it is precisely the figure of the spy that is the hauntological element, "both visible and invisible, both phenomenal and nonphenomenal" (Derrida 2002: 117) - a "sign of trauma and its mourning" which allows us, "in an important way", to do "the work of history" (Freccero 2013: 338). That is, the hauntological acts as what Foucault would call a new type of regularity, which is to say a historical a priori - an invisibilised structuring structure. This is especially important for understanding algorithmic governmentality and its reliance on surveillance, as I will argue in the section on algopopulism.

Arendt's invaluable account aside, it remains what Gilles Deleuze and Félix Guattari would describe as the macropolitical. For them, however, fascism cannot only be explained in terms of national-historical phenomena. Instead of trying to describe fascism in the way Arendt did - which, to be sure, provided an important framework for understanding totalitarianism and fascism - they are interested in the molecules of fascism all around us, how fascism "survived its historical mutations and adapted itself to contemporary institutions" (Genosko 2017: 59). According to them, then, fascism is a "cancerous body" rather than a "totalitarian organism" (Deleuze and Guattari 1987: 215) - something "in us all, in our own heads, and in our everyday behaviour" that "causes us to love power" and "desire the very thing that dominates and exploits us", as Foucault writes in his preface to Anti-Oedipus (Deleuze and Guattari 1983: xiii). Having taken this view, Deleuze and Guattari set out to account for what they call microfascism: "microformations" that shape "postures, attitudes, perceptions, expectations, semiotic systems, etc." (Deleuze and Guattari 1987: 215).


Microfascisms: accounting for desire and affect with Deleuze and Guattari

Why did people vote for Hitler? Why did they vote for Trump? Or Bolsonaro for that matter? Were they duped? Was it some kind of mass hysteria or mass psychosis? Deleuze and Guattari, following Wilhelm Reich would say no, at some point the people desired fascism. The question they grappled with accordingly and wanted to answer is: What causes people to desire their own repression -that which is against their own interests? And how do we theorise and understand the "investments of desire that shape power, and diffuse it, such that power is located as much at the level of a cop as that of a prime minister"? (Deleuze 2004a: 212). That is to say, instead of understanding fascism as a break in history and trying to provide us with a new framework for understanding it in national-historical terms, they view fascism as a modification at the level of desire which causes people to desire a specific distribution of power, specifically negative power or power over (pouvoir) rather than power to (puissance), the latter of which expresses positively potentiated power, for example the power to create, the power to effect social change and the power to love.

What becomes immediately clear here is that Deleuze and Guattari do not consider fascism only as generalisable in terms of an ideology, or a "particular arrangement of the state", or even of a particular arrangement of the "state to its population" (May 2013: 19). They want, rather, to understand even ideology in terms of desire - that vital force which "constitutes the very texture of society in its totality, including its mechanisms of reproduction" and repression (Deleuze 2004b: 266). To do so, they have to start by unshackling desire from some of its earlier interpretations, such as lack vis-à-vis Lacan, or the Oedipal master signifier vis-à-vis Freud. They do this by taking umbrage, in Anti-Oedipus, at the way in which psychoanalysis reduces every form of desire - an entire libidinal economy - to a specific familial formation that "closes the familial triangle over the entire unconscious" so that it no longer "has anything to do with the social field actually invested by the libido" (Deleuze and Guattari 1983: 55, 62). Their argument, on the contrary, is that desire is invested in or directed and arrested by myriad social arrangements - be that family, friends, politics, religion, capitalism or popular culture. More to the point, they hold that desire produces the social field. So instead of packaging desire according to "readymade encodings", such as 'lack' or 'Oedipus', desire on their view "tends to wander" and evade prefabricated meanings (Genosko 2017: 61) - which is also how and why desire can become subverted or 'sick' and come to desire its own repression. And it is precisely this subverted or inverted desire that they link to microfascism - a tendency that exists in all of us, in all our heads. I would argue, therefore, that whereas a generalised espionage is the hauntological dimension of national-historical phenomena, microfascism is the spectral dimension of desire because it constitutes those latent historical a priori - like racism and sexism - inside our own minds that can become weaponised by politicians or, in the case of the digital milieu, by social media. Think here, for example, of how Trump "gave voice to the sense of frustration and fear felt by the traditional American male nostalgic for his old gender privileges and alarmed by the seemingly unstoppable rise of 'women's power'" (Revelli 2017: 77), or "how the reaction to immigration" became a "driver of the Leave vote" that led to Brexit in the UK (Revelli 2017: 95). For political scientist Marco Revelli, these examples illustrate why populism is different from other 'isms' such as "socialism, communism, liberalism, fascism and so on"; because rather than being identifiable in terms of a relatively stable label with relatively stable characteristics, it "is a mood" (Revelli 2017: 10-11). To rephrase it, there is an I feel at the heart of fascism - and it is this affective or desirous dimension, along with surveillance, that form two of the most salient vectors in the algorithmically aided populisms we have been witnessing in recent history, accelerated by AI's tendency "to propagate patterns of carelessness and extractiveness alongside a concentration and centralization of power" (McQuillan 2022: 10). But while there are historical correlations here to the theorisations of Arendt and Deleuze and Guattari, we would do well to remember that fascism is not a solvable problem - that would be to suppose that it has ready-made answers. This, precisely, is why Brad Evans and Julien Reid argue that fascism is a problematic which needs to be continually reproblematised - "so that the ways in which we respond to it can do other forms of work in the world than simply reproduce the old formulae" (Evans and Reid 2013: 3). In other words, failing to reproblematise fascism not only pre-empts the types of solutions that are imaginable, but also sets up the very conditions or limits of the imaginable. New symptomatologies of fascism, conversely, render it open to new understandings, but only if we start any analysis with an acceptance "that we are all, always, fascists of multiple kinds" (Evans and Reid 2013: 5), thus recognising that it is our ethical task to expose our spectral selves to ourselves.

In the following section, then, I draw on Arendt's and Deleuze and Guattari's theorisations to make clear the historical correlatives between the spy and contemporary surveillance on the one hand, and desire and the harnessing of affect through machine learning methods and their deployment via social media platforms on the other. At the same time, I attempt to reformulate fascism in terms of the new populisms that undergird it, as well as the role of AI and the "technological shift in the framework of society" (McQuillan 2022: 1) that it has induced, resulting in "social re-engineering" marked by "abstraction, distancing and optimization", and the "reproduction of scarcity" via the logics of austerity (McQuillan 2022: 74) - all of which create fertile soil for the fostering, proliferation and spreading of authoritarian politics.


Algopopulism: algorithmic governance, recursive conduct and a generalised proletarianisation

Algorithms have become so ubiquitous that they could be said to have both geopolitical and philosophical implications (Bratton 2015: 69). "From the rationalization of labor and social relations to the financial sector, algorithms are grounding a new mode of thought and control" (Parisi 2015: 125) that has been efficacious for the pervasive and accelerated spread of populist and authoritarian politics. But long before algorithms and social media platforms were utilised by politicians, the former Italian Prime Minister and media tycoon, Silvio Berlusconi, recognised the power of technology - in his case, the television - to "change the relationship between leader and people" (Revelli 2017: 172). Moreover, this predecessor of social media politicians understood the language of classical populism: giving "the image of a man who came from the outside"; creating a chasm between him and "those 'up above', in the Palaces of Power"; and using a tone of urgency to reinforce the idea that politics, Italy and 'the people' need him (Revelli 2017: 172). With the decline of TV, however, came the inevitable waning of Berlusconism and, with it, video-politics and tele-populism. That is, until Beppe Grillo1, one of the founders of the Five Star Movement - a political party in Italy - defeated the Democrats in a landslide victory. Interestingly, it was the "emergence of the web as a new, all-encompassing media universe" that was the "defining factor in the parallel emergence of grillismo and its innovative 'cyber-populism'" (Revelli 2017: 179). The point is that while tools afforded to politicians

- such as those of Cambridge Analytica - may seem entirely unprecedented, there is a longer history between technology and populism, as we see here. What makes the contemporary use of technology different, however, is, in the first place, its surveillance capabilities which far outweigh those of a generalised espionage, even though there are clear historical correlates and spectral remnants between algorithmic surveillance and the figure of the spy, as well as its close cousins, panopticism, as described by Foucault2, and the societies of control, as outlined by Deleuze3. As is the case of disciplinary and control societies, which are premised on a certain kind of logics - panopticism and modulation respectively

- the current generalised environment of ubiquitous data surveillance too is premised on a particular kind of logic: that of extraction and optimisation through statistical monitoring which "has deep Cold War roots" (McQuillan 2022: 15). These surveillance techniques are, moreover, used to harness and weaponise affect - and this is the second aspect that distinguishes contemporary technologies from older ones. Think, for example, of the speed and scale of contemporary technologies and what becomes possible when it combines with the surveilled mobilisation and weaponisation of nationalist, xenophobic, sexist and racist affects for populist ends through algorithmic governmentality - "the idea of a government of the social world that is based on the algorithmic processing of big data sets rather than on politics, law, and social norms" (Rouvroy 2020). Here Antoinette Rouvroy is, of course, extending Foucault's understanding of government and governmentality which he sees as entangled not only with the subject, but also with power. In fact, Foucault starts his essay on "The Subject and Power" by talking about the two "pathological forms" or "diseases of power", namely "fascism and Stalinism" (1982: 779). Recall that, for Foucault, government is understood to refer not only to "political structures or to the management of states", but also to the calculated ways in which it conducts (directs) the conduct (behaviour and habits) of people, thus either facilitating or foreclosing the possibilities of their thought and action in advance (1982: 790). This is why Foucault describes governmentality as the "conduct of conduct" (Foucault 2008: 186). A different way of thinking about this is as the governmentalisation of power relations.

To get a clearer idea of what it means to extend governmentality to algorithmic ecologies, it is useful to come to grips with what exactly 'algorithmic' means here. Simply put, algorithmic refers in this context to what is sometimes synonymously called AI (artificial intelligence) - "a cross-disciplinary approach to understanding, modeling, and replicating intelligence and cognitive processes by invoking various computational, mathematical, logical, mechanical, and even biological principles and devices" (Frankish and Ramsey 2014: 1). AI - or machine learning methods, such as deep learning - differs significantly from older approaches based on the manipulation of symbols by means of explicit rules or logic, but also because "instead of a programmer specifying the sequence of operations which produce the desired result, machine learning algorithms are fed a sample of the required results and use statistical estimation to figure out how to reproduce them" (McQuillan 2022: 11). In other words, machine learning improves a programme's predictive performance through iteration and optimisation. As it relates to governmentality, then, we can say that individuals are managed as data, which involves bypassing "subjectivity by automation", so reducing the subject to a "collection of infra-individual data" that are "recomposed at a supra-individual level under the form of a profile" in a kind of pure, recursive totality (Rouvroy and Stiegler 2016: 11, 12). Thus, I would argue, we can think of algorithmic governmentality as effectuating recursive conduct. So because the art of government, by which people are rendered governable, has undergone a digital shift, we find not only that people are treated as data and the social world as a problem of big data sets, but also that there emerges from this form of governmentality a novel and recursive rationalisation, elaboration and centralisation - or exercise - of power and power relations. This is why Dan McQuillan argues that "AI is always more than a set of machine learning methods" - because we cannot "separate the calculations in the code from the social contexts of its application", nor can we divorce it from "the assembly of institutional arrangements" needed for it to be deployed impactfully in society (2022: 1).

This shift towards algorithmic governmentality was enabled by the convergence of five major developments: the proliferation of data; the development of deep convolutional networks that use highly optimised algorithms to analyse the extracted data; networks that allow for the immediate and economical transmission of data and their analyses; spacious and flexible storage via the cloud - "the governing nexus" or nomos of information (Bratton 2015: 111); and significant improvements in hardware that "have added sheer power to the capture and analysis of data" (Kalpokas 2019: 2-3). It was from this convergence, then, that the cybernetic episteme emerged, characterised by digitised modulation, and marked by recursive feedback loops that prompt continuous adaptation to both "the working principle of computation" and a novel way of thinking that "extends beyond computers to social and economic model ing" (Hui 2018: 147). This shift, according to Bernard Stiegler, is not merely technical, which is to say a transition from analogue to digital, but organological

- a question of life, where the biological has become technical (in as much as the technical has become biological), structured by the organic and the inorganic alike. This is important for thinking about new forms of populism and fascism because, while algorithms "involve the performance of mathematical functions, at least at the level of the machinic code", they also actively "imagine" and approximate "courses of action associated with things or states of the world" (Amoore and Piotukh 2016: 4). In effect, this means that algorithms are imbricated in processes of worldmaking with and for us. We know, for example, that algorithms filter what we see and don't see, creating "novel ways of perceiving the world" through "new visibilities and invisibilities" (Amoore and Piotukh 2016: 5). This, in turn, provokes questions about the interests that "emerge from such calculative processes" (Amoore and Piotukh 2016: 5). What, for example, happens to your interests when what you see is filtered to amplify certain feelings of insecurity? What happens to knowledge when it becomes automated and recursive?

These are important questions for Stiegler who argues that knowledge is first and foremost the "knowledge of savoir-faire (know-how, skill) and savoir-vivre (knowledge of how to live)" (2013: 3). According to him, however, knowledge production has been disrupted by "the industrialization of knowledge" and "cognitive technologies - in particular insofar as these produce a syndrome of cognitive saturation, and have entropic effects" on our ways of knowing how to live and dream, which he refers to as a generalised proletarianisation (Stiegler 2013: 3). Here, Stiegler is extending Marx's notion of proletarianisation, which initially referred to the creation and expansion of the working class, or proletariat, who, instead of owning the means of production, were required to sell their labour-power to the bourgeoisie - the class of people who did own the means of production and acquired their wealth from "the extortion of surplus-value from the workers" (Marx 1976: 33). In Stiegler's work, we find three forms of proletarianisation: first, of the producer; second, of the consumer; and third, a generalised proletarianism. The first form of proletarianisation is drawn "directly from Marx" who sees the "worker's knowledge" as having been "inscribed in the machine" (Dillet 2017: 82). In other words, during the first industrial revolution, marked by the "invention of the steam engine and mechanized production" (Dillet 2017: 82), an exosomatisation - or externalisation of memory and knowledge

- occurs. That is, workers transfer a certain amount of their knowledge to machines. The second industrial revolution, which can be described as Taylorism-Fordism, provoked yet a further exosomatisation: this time signalling a loss of knowledge through consumerism. Put simply, knowledge becomes subsumed by the culture industry which both stimulates an appetite for amusement and cultivates a false sense of security - material and psychological - vis-à-vis consumption (see Horkheimer and Adorno 1989). Finally, we get to our current situation which, through "the financialization of society and debt" on the one hand (Dillet 2017: 82), and digitalisation on the other, induced a generalised proletarianism, encompassing the loss of work-knowledge (savoir-faire) through mechanisation; the loss of life-knowledge (savoir-vivre) through consumption; and the loss of conceptual knowledge due to the continual and near-automatic junking of knowledge and memory into devices like cell phones (Stiegler 2006; 2019: 24). For Stiegler, this latter industrial revolution is characterised by "digital reticularity" which has "led to the destruction of the long-term temporality specific to politics" (Dillet 2017: 99). This loss of knowledge is precisely what creates the right conditions for populisms to thrive, especially when algorithms become involved, because algorithmic reason incites the unbinding of the drives through automation, thus reticulating noetic life and inducing a loss of reason, as well as reasons for living and dreaming.

The problem is not as simple as exosomatisation or algorithms, of course. We have, at any rate, always been technical, forever exteriorising memory and knowledge partially and mnemotechnically - "whether via tools and artefacts or social formations, organizations and rituals" (Gray 2022: 89). The reason for this is that such "exteriorized memory, also known as hypomnesis, frees anamnesis or functional memory from its overall dependence on the human mind by becoming located in a technical object, the form of exteriorization" (Gray 2022: 89). So the problem is not exosomatisation per se, but what happens when it is driven by algorithms and the assemblages they form part of, which includes aspects such as the concentration of power via data collection, information hoarding and use; the emergence and dissemination of fake news, misinformation and disinformation and the effects of these, such as political polarisation, right-wing mobilisation and election hacking; inscrutable surveillance practices infused by economic imperatives; technological over-dependence and addiction; and a sharp increase in mental illness, especially in younger generations through the exploitation of available brain time. It is this combination of factors that drives the generalised proletarianisation of our age, "systematically targeting drive-based and regressive mechanisms" which operate "against spirit", where spirit implies "the care taken of the objects and subjects of individual and collective desire" (Stiegler 2013: 6). This loss of spirit, which is to say care, along with the loss of knowledge, is precisely what underlies what I have termed algopopulism - the new kinds of algorithmically amplified fascist politics animated by "cynicism and a transgressive acting out" that are themselves informed "by a pure death drive" (Stielger 2013: 6). As a result, this form of politics "fails to lead to any transindividuation, or to any possibility of sublimation", excepting negative sublimation (Stielger 2013: 6). The consequences are disturbing, to say the least, because this kind of negative sublimation, enabled by continuous surveillance and data extraction, and driven by algorithmically iterated, optimised and weaponised affect, disrupts our ability to produce shared horizons through social practices of care and collective meaning-making that can carry futural dreams across the passage of time. Instead, a "pathogenic form of sublimation" is elicited (Stiegler 2013: 60), giving rise to a paralysis of the human spirit so that both psychic (individual) processes of individuation and collective processes of transindividuation that exceed "the I and connects it to the we" are irrevocably disrupted (Stiegler 2013: 2). Thus disrupted, there "is no longer any 'we'; there is only the 'they'" (Stiegler 2013: 86). This transformation, I would argue, lies at the heart of algopopulism which epitomises an algorithmically fuelled politics of the 'they' - a spiritual misery that leads "to the outright destruction of societies, to their death" (Stiegler 2013: 125).


But is it fascism?

I recently watched Louis Theroux's new three-part documentary series called Forbidden America (2022). In the opening episode, titled "Extreme and Online", he interviews a number of self-proclaimed right-wing 'rising stars', such as Nick Fuentes, Beardson Beardly and Baked Alaska. For those unfamiliar with these names, Nick Fuentes - whose views are so far right that he has been banned from Republican Party gatherings - is the head of the white nationalist America First movement and founder of AFPAC, or the America First Political Action Conference, a far-right alternative to CPAC, the Conservative Political Action Conference. The third AFPAC conference was famously attended by a number of US elected officials, including Marjorie Taylor Greene. Beardson Beardly, known as one of Nick's 'Generals', is interviewed in the documentary by Theroux, specifically about his Nazi salute at one of the AFPAC events. Despite his obvious involvement in the group and clear racial separatist views, Beardson Beardly is quick to dismiss any suggestion that he is a white nationalist when meeting Theroux, even though he concedes that the movement is about race. In fact, he says to Theroux, he does not even consider himself political - at core he's a gamer, a punk-rock kid who stands for anti-authoritarianism. And anyway, he adds, it was absolutely not a Nazi salute (though anyone with eyes to see would surely doubt this). When Theroux interviews Baked Alaska (a.k.a. Tim Gionet), he goes as far as to exclaim that we need more racist jokes. Watching the interviews with these young men - yes, Theroux interviews mostly white men - one gets the sense that they are all "guided by a powerful sense that America has been lost in some way" which, we know, is "a classic form of nationalist nostalgia", often couched "in irony and humour, both as a defense when criticised and to disguise anything more sinister" (Thompson 2022). But is it fascism?

The simple answer to this is no, though it can be argued that this episode of Theroux's documentary and the second vignette - the storming of the US Capitol

- represent good examples of how algopopulism works. And, while they do not resemble the historical presidents set by Mussolini or Hitler, it is clear that certain microfascistic leanings and their attendant affective states have successfully been exploited and propelled via algorithmic surveillance - which includes here the personalisation of social media feeds. It is these personalised feeds, characteristic of algorithmic governmentality, that produce recursive conduct - the exercise of power and power relations via a recursive governmentality that disrupts the kind of long-term knowledge needed for resistance. What would be needed for resistance, then, following Stiegler, is a deproletarianisation, or the production of new knowledges - and specifically knowledges that can be transferred across generations - through a politics of care.

As for the remaining two vignettes, I would venture here to say that the third one, which deals with Agamben's sentiments around the pandemic, was, at least in part, driven by recursive conduct and serves, perhaps, as an admonishment to us all. The first vignette, too, clearly shows how recursive conduct becomes normalised, even though it is not an example of algopopulism per se. Moreover, it reveals something about the conditions that give rise to populism and fascism. Specifically, it illustrates something about what Stiegler calls adaptation, which he opposes to adoption. For Stieger, adaptation is not a maladjustment, but a "hyperadaptation" to a situation - in this case, digitality

- and, as he warns, "recourse to adaptation feeds political conservatism", all the while establishing the idea that "there is no alternative" (Stiegler n.d.). This kind of digitised adaptation, characterised by recursive feedback loops, prompts processes of disindividuation: the inability to individuate - both personally and socially - because of a premature disruption of processes of subjectivation which leads to an inability to create long-term social memories and practices that can resist foreclosing future possibilities. Adoption, on the other hand, produces an "enrichment" rather than "a restriction of the possibilities of the individual" (Stiegler n.d.). As such, it constitutes the process of individuation, grounded in an economy of care, which is to say the resistance of carelessness promulgated by algopopulism and recursive conduct.


An uncanny conclusion

What would it mean to 'commune with' algorithms, such as those associated with the propagation of fake news, misinformation and disinformation aimed at political division, right-wing mobilisation and election hacking? What would it mean to sit with these ghostly apparitions that haunt our lives so with their presence/absence coordinates? Might a hauntological approach to algorithms draw from us an ethos of attunement to "manners, customs, shared experiences and memories, ambiences, atmospheres, displaced narratives, structures of feeling" (Blackman 2019: 25) and other kinds of seen/unseen categories that shape our knowledges of how to live and dream together? Is this not what is required for symptomatology to become fabulation, to become resistance? To animate what is through a coming back to that is also a making known, a laying bare that which is absent/present. And how might we consider the 'afterlives' of algorithms? Could a hauntological approach to them direct "our attention to what already exists in the present as a form of anticipation, to a time in the present that has not passed, and which operates as an attractor for possible futures" (Blackman 2019: 56)? Who's to say that the afterlives of algorithms can't "attract, collect and channel fragments, gaps, anomalies, puzzles and parallel temporalities", thereby "producing the potential for 'queer aggregations'" (Blackman 2019: 56)? Here, in the ruins of the infra-world, in the ruins of our world, do we not owe it to ourselves - and to the ghosts that haunt our lives - to become a kind of medium? No, "to make our own medium" (Tuck and Ree 2013: 650). To undertake with our spectral companions a clairvoyant journey to the very heart of the disruption that Stiegler conceives of as a generalised proleratianisation: the loss of work-knowledge (savoir-faire), life-knowledge (savoir-vivre) and conceptual knowledge. And is this, moreover, not the work that counters the fascism in us all, in all our heads?



Agamben G. 2020. The invention of an epidemic (column entry). Quodlibet. 26 February. Available at: [accessed on 23 November 2022].         [ Links ]

Agamben G. 2021. Second-class citizens (column entry). Quodlibet, 16 July. Available at: [accessed on 23 November 2022].         [ Links ]

Ahmed S. 2019. The messy truth about social credit. Logic. 1 May. Available at: [accessed on 23 November 2022].         [ Links ]

Amoore L and Piotükh V. 2016. Algorithmic life: calculative devices in the age of big data. London, New York: Routledge.        [ Links ]

Arendt H. 1962. The origins of totalitarianism. Cleveland, New York: The World Publishing Company.         [ Links ]

Argentino M-A. 2021. QAnon and the storm of the U.S. Capitol: the offline effect of online conspiracy theories. The Conversation. 7 January. Available at: [accessed on 23 November 2022].         [ Links ]

Blackman L. 2019. Haunted data affect, transmedia, weird science. London, New York: Bloomsbury.        [ Links ]

Bratton B. 2015. Outlining artificial intelligence: reckoning with turing tests. In: Pasquinelli M (ed). Alleys of your mind: augmented intelligence and its traumas. Lüneberg: Meson Press.         [ Links ]

Bratton B. 2021. Agamben WTF, or how philosophy failed the pandemic (blog). Verso. 28 July. Available at: [accessed on 23 November 2022].         [ Links ]

Davis C. 2005. État présent: hauntology, spectres, and phantoms. French Studies LIX(3): 373-379.        [ Links ]

Deleuze G. 1988. Foucault. Transl S Hand. London: Athlone Press.         [ Links ]

Deleuze G. 1992. Postscript on the societies of control. October 59: 3-7.         [ Links ]

Deleuze G. 2004a. Intellectuals and power. In: Lapoujade D (ed). Desert islands and other texts 1953-1974. Transl M Taormina. London, Cambridge MA: MIT Press.         [ Links ]

Deleuze G. 2004b. On capitalism and desire. In: Lapoujade D (ed). Desert islands and other texts 1953-1974. Transl M Taormina. London, Cambridge MA: MIT Press.         [ Links ]

Deleuze G. 2007. The brain is the screen. In: Lapoujade D (ed). Two regimes of madness: texts and interviews 1975-1995. Transl A Hodges and M Taormina. Los Angeles, CA: Semiotext(e).         [ Links ]

Deleuze G and Guattari F. 1983. Anti-Oedipus: capitalism and schizophrenia Vol. 1. Transl R Hurley, M Seem and HR Lane. London, Minneapolis: University of Minnesota Press.         [ Links ]

Deleuze G and Guattari F. 1987. A thousand plateaus: capitalism and schizophrenia Vol. 2. Transl B Massumi. London, Minneapolis: University of Minnesota Press.         [ Links ]

d'Entrèves MP. 1993. The political philosophy of Hannah Arendt. London, New York: Routledge.         [ Links ]

Derrida J. 2002. Spectrographies. In: Derrida J and Stiegler B (eds). Echographies of television: filmed interviews. Cambridge: Polity Press.         [ Links ]

Dillet B. 2017. Proletarianization, deproletarianization, and the rise of the amateur. Boundary 2 44(1): 79-105.        [ Links ]

Evans B and Reid J. 2013. Introduction: fascism in all its forms. In: Evans B and Reid J (eds). Deleuze and fascism. security: war: aesthetics. London, New York: Routledge.        [ Links ]

Foucault M. 1982. The subject and power. Critical Inquiry 8(4): 777-795.        [ Links ]

Foucault M. 2008. The birth of biopolitics: lectures at the Collège de France, 1978-79. Transl G Burchell. New York: Palgrave Macmillan.         [ Links ]

Frank ish K and Ramsey WM. 2014. Introduction. In: Frankish K and Ramsey WM (eds). The Cambridge handbook of artificial intelligence. Cambridge: Cambridge University Press.        [ Links ]

Freccero C. 2013. Queer spectrality: haunting the past. In: Pilar Blanco M (del) and Peeren E (eds). The spectralities reader: ghosts and haunting in contemporary cultural theory. London, New York: Bloomsbury Press.         [ Links ]

Genosko G. 2017. Black holes of politics: resonances of microfascism. La Deleuziana 5: 59-67.         [ Links ]

Gray C. 2022. Anarchism after Deleuze and Guattari: fabulating futures. London, New York: Bloomsbury.        [ Links ]

Gray C and Eloff A. 2022. Fabulation in a time of algorithmic ecology: making the future possible again. In: Das S and Pratihar AR (eds). Technology, urban space and the networked community. London: Springer Nature.        [ Links ]

Horkheimer M and Adorno TW. 1989. Dialectic of enlightenment. Transl J Cumming. New York: Continuum.         [ Links ]

Hui Y. 2018. Archives of the future: remarks on the concept of tertiary protention. In: Inscription. Göteborg: Landsarkivet i Göteborg, 129-151.         [ Links ]

Kalpokas I. 2019. Algorithmic governance: politics and law in the post-human era. Switzerland: Palgrave Macmillan.        [ Links ]

Loucaides D. 2019. What happens when techno-utopians actually run a country. Wired. 14 February. Available at: [accessed on 23 November 2022].         [ Links ]

Longley R. 2021. Totalitarianism, authoritarianism, and fascism. ThoughtCo. 6 December. Available at: [accessed on 23 November 2022].         [ Links ]

Marx K. 1976. Capital Volume 1. Transl B Fowkes. London: Penguin Books.         [ Links ]

McQuillan D. 2022. Resisting AI: an anti-fascist approach to artificial intelligence. Bristol: Bristol University Press.        [ Links ]

Parisi L. 2015. Instrumental reason, algorithmic capitalism, and the incomputable. In: Pasquinelli M (ed). Alleys of your mind: augmented intelligence and its traumas. Lüneberg: Meson Press.         [ Links ]

Pilar Blanco M (del) and Peeren E. 2013. Haunted historiographies/introduction. In: Pilar Blanco M (del) and Peeren E (eds). The spectralities reader: ghosts and haunting in contemporary cultural theory. London, New York: Bloomsbury Press.         [ Links ]

Revelli M. 2017. The new populism: democracy stares into the abyss. Transl D Broder. London, New York: Verso.         [ Links ]

Rouvroy A. 2011. Algorithmic governmentality and the death of politics. Green European Journal. 27March. Available at: [accessed on 23 November 2022].         [ Links ]

Rouvroy A and Stiegler B. 2016. The digital regime of truth: from the algorithmic governmentality to a new rule of law. Transl A Nony and B Dillet. La Deleuziana 3: 6-29.         [ Links ]

Rouvroy A. 2020. Algorithmic governmentality and the death of politics. Green European Journal, 27March. Available at: [accessed on 23 November 2022].         [ Links ]

Stiegler B. 2006. Anamnesis and hypomnesis: the memories of desire. Ars Industrialis. 8 April. Available at: [accessed on 23 November 2022].         [ Links ]

Stiegler B. 2013. Uncontrollable societies of disaffected individuals. Transl D Ross. Cambridge: Polity Press.        [ Links ]

Stiegler B. 2019. The age of disruption: technology and madness in computational capitalism. Transl D Ross. Cambridge: Polity Press.         [ Links ]

Stiegler B. n.d.. Adaptation-adoption. Ars Industrialis. Available at: [accessed on 23 November 2022].         [ Links ]

Tazzioli M and Pârvan O. 2021. Technologies of control and infrastructures of redistribution. e-flux 123: 1-12.        [ Links ]

Thompson J. 2022. Louis Theroux's Forbidden America review: a sinister warning about the Far Right. Evening Standard. 14 February. Available at: [accessed on 23 November 2022].         [ Links ]

Tuck E and Ree C. 2013. A glossary of haunting. In: Holman JS, Adams TE and Ellis C (eds). Handbook of autoethnography. Walnut Creek, CA: Left Coast Press, Inc.         [ Links ]



First submission: 16 March 2022
Acceptance: 24 November 2022
Published: 31 December 2022



1 While I do not have space here to discuss the creation of grillismo, I highly recommend the 2019 article, "What Happens When Techno-Utopians Actually Run a Country" by Darren Loucaides, published in Wired. The article explains how it was really Gianroberto Casaleggio, a software designer who worked at Olivetti for a number of years, and co-founder of the Five Stars, who understood the power of technology for the movement, not Grillo himself.
2 See: Foucault, M. 1977. Discipline and Punish: The Birth of the Prison. Transl A Sheridan). London: Penguin.
3 See: Deleuze G. 1992. Postscript on the Societies of Control. October 59: 3-7.

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License