We have entered the post-alphabetic era announced by Vilém Flusser. Apparatuses now program us as much as we program them. In this techno-imaginal universe where all our faculties are traversed by technologies, the question of fascism arises differently. Vectofascism designates this contemporary political form that adapts the fundamental mechanisms of historical fascism to the technological, communicational, and social structures of the digital era. Trump’s second term is not a simple “return” to fascism, as if history followed a circular schema, but a fractured re-emergence in a radically distinct techno-affective field. These movements can be observed in different geographical contexts, with variations: Erdoğan in Turkey, Orbán in Hungary, Farage in the United Kingdom, Meloni in Italy, Putin in Russia, Le Pen and Zemmour in France. The prefix ‘vecto-‘ indicates precisely this transformation: we are no longer in a politics of masses but in a politics of vectors, of lines of force and directional intensities that traverse and constitute the algorithmized social body. Even speaking of ‘masses’ would still be nostalgic, a residual concept from an era when physical density manifested the political. Today, density is expressed in terms of aggregated attention, of micro-impulses algorithmically synchronized in the absence of any bodily proximity. Bodies no longer need to touch each other to form a political force; it suffices that their data touch each other in the latent space of servers.#Vilem Flusser, #tech-fascism
Vectorization as Onto-Political Process
Before examining the political implications of vectorization, it is essential to understand precisely what vectors are in the context of contemporary AI systems. In machine learning, a vector is a mathematical representation: a sequence of numbers arranged in n-dimensional space. For example, the word ‘democracy’ might be represented as a point in a 4096-dimensional space, where each dimension captures a statistical correlation with other terms in the training corpus. This representation is fundamentally different from symbolic or semantic encoding. The vector does not contain a definition of ‘democracy’ in any human-readable sense. Instead, it encodes patterns of co-occurrence: which words appear near ‘democracy’ in texts, in what contexts, with what frequency. The meaning is distributed across all dimensions simultaneously, with no single dimension corresponding to a comprehensible feature like ‘political system’ or ‘freedom.’ Crucially, these latent spaces have mathematical properties that enable computation. Vectors can be added, subtracted, and compared. The famous example ‘king – man + woman ≈ queen’ demonstrates how semantic relationships can be captured as geometric relationships in vector space. Proximity between vectors indicates statistical similarity in how terms are used, not necessarily semantic or conceptual similarity in human terms.
This is what makes latent space ‘latent’: it is hidden, not directly observable or interpretable. It is a compressed representation that captures patterns in data without making those patterns explicitly legible to human understanding. The space is operational—it allows algorithms to make predictions, generate content, classify inputs—without being transparent. The vectorial de-semanticization transforms information itself into algorithmic currency. Information no longer circulates as semantic use-value, but as computational exchange-value. Today, vectorization constitutes a fundamental onto-political process that structures the very texture of the contemporary social fabric. It is the process by which social entities—individuals, groups, communities—are transformed into carriers of directional variables, that is, into vectors endowed with a predetermined orientation in a conceptual space saturated with differential values. The critical shift occurs when this technical process of representing information as vectors becomes applied to human beings and social categories. When an individual is designated by the signifier ‘foreigner,’ this designation does not simply constitute the attribution of a descriptive property, but the immediate inscription in an oriented force field—a vectorization that operates analogously to how AI systems encode words. Just as ‘democracy’ becomes a point in abstract mathematical space, the ‘foreigner’ becomes a position in social space defined by multiple dimensions: legal status, economic position, cultural markers, geographic origin. Similarily, in digital social networks, each profile becomes the point of application of a multiplicity of assignative vectors: gender, age, origin, socio-economic position. Each interaction constitutes a moment of this generalized vectorization that transforms singular existence into a series of abstract coordinates.
The parallel is not merely metaphorical. The same mathematical frameworks used in AI—neural networks, embedding spaces, clustering algorithms—are directly applied to social data. Recommendation systems, credit scoring algorithms, predictive policing tools, and targeted advertising platforms all operate by representing people as vectors in multidimensional spaces and then performing computations on those representations. While historical fascism still operated in three-dimensional Euclidean space; vectofascism operates in an n-dimensional hyperspace where the very notion of ‘gathering’ has changed dramatically: What gathers are no longer bodies in a stadium but data in a vector space. Traditional political categories (right/left, progressive/conservative) become impoverished projections of a complex multidimensional space manipulated by algorithms. Uniform crowds operated in Euclidean space; vectofascist alignment operates in a latent space that escapes our representation. What gathers is no longer carried by self-consciousnesses but by statistical resemblances. This analysis of vectofascism is inscribed in a theoretical genealogy that should be specified. McKenzie Wark, in A Hacker Manifesto (2004), identified the emergence of a ‘vectoralist class’: this new capitalist formation that appropriates access to communication vectors and alone possesses the means to realize the value of informational creations. “The new class struggles over information property now oppose the hacker class, which creates, invents, explores, discovers, to the vectoralist class” (ibid.). This formulation, elaborated in the early 2000s, anticipated with remarkable prescience the transformations we observe today. Wark diagnosed the emergence of a new capitalist formation; the concept of vectofascism extends this analysis into a different historical conjuncture by observing its crystallization into a specific political regime (For the difference between vector classes and vectofascism see: https://chatonsky.net/classes-fascisme).#patterns, #LLM, #latent space
Still, she maintains the possibility of a critique external to the relations she analyzes; she can identify the vectoralist class because she presupposes a viewpoint that escapes it. This class according to Wark, “appropriates free information and transforms it into private property,” thus creating an objective antagonism with the “hacker class” (ibid.). Vectofascism operates precisely through the saturation of this exteriority. It functions in an n-dimensional latent space that cannot even be directly visualized by the human mind. How can ‘hackers’ maintain their common position in a regime where the very categories by which they identify themselves—’creativity,’ ‘innovation,’ ‘informational freedom’—are decomposed into algorithmically manipulable mathematical vectors? This distinction does not render Wark’s analysis obsolete. On the contrary: it reveals a temporal stratification of critique. Wark identifies the logic of accumulation that persists under technological transformations. ‘Vectofascism’ describes how this logic reconfigures itself according to an algorithmic governmentality that transforms value extraction: it no longer bears on abstract labor, but on life itself in its most intimate dimension. Each gesture, each emotion, each relationship becomes exploitable data. We are no longer facing an exploitation of labor in its traditional sense, but a modulation of what constitutes ‘labor’ itself to encompass attention, affect, and subjectivity, that is to say, existence.#vectorial class, #neoliberalism, #algorithm
Vectofascism retains three fundamental characteristics of historical fascism, while adapting them to contemporary conditions:
1.) The cult of the vectorial leader
In the universe of technical images, the leader is no longer a subject carrying a historical will but a function in a feedback system. The cult of the vectofascist leader is no longer a cult of the person but a cult of the interface, of the interaction surface. The leader has no inner life; he is merely a projection surface whose semantic content responds to the intensity of external projections. This is why it is so difficult to discuss the content of his speeches rationally. What is adored is not the supposed depth of the leader but their capacity to function as a perfect projection surface. Grotesquerie becomes an essential political operator. The body of the vectofascist leader can free itself from the requirement of perfection precisely because it no longer has to incarnate but to channel. More precisely, vectofascist leaders do not simply inhabit this interface passively, they actively react to and calibrate themselves according to vectorial probabilities. Their campaigns are strategically attuned to electoral segments algorithmically identified and targeted through platforms like Facebook, Instagram, and TikTok. Each policy statement, rhetorical gesture, and media appearance is micro-optimized for maximal engagement across specific demographic vectors. The leader becomes a kind of intelligent interface that learns from real-time feedback loops: which messages generate the strongest emotional responses in which populations, which visual frames trigger the highest engagement metrics, which talking points resonate most powerfully within particular affective clusters.#tech-fascism, #leadership, #platform
This is not propaganda in the traditional sense (directed toward a mass ‘people’) but a form of algorithmic clientelism: each voter segment receives a customized version of the leader’s persona and messaging, calculated to produce maximal affective resonance for that specific vector. The manifestly constructed, artificial, even ridiculous character of appearance is not a defect but an asset: it signals that we have fully entered the regime of the technical image, where the referent fades behind its own representation. This argument extends to European vectofascism. Viktor Orbán’s strategic oscillation between democratic legitimacy and nationalist defiance, Giorgia Meloni’s careful repositioning of fascist aesthetics within European respectability, and Marine Le Pen’s calibrated nationalist rhetoric all function identically: as adaptive interfaces testing which visual codes and rhetorical formulations maximize engagement within their specific networks. The deliberately artificial elements of their presentation, whether Trump’s appearance or Meloni’s symbolic transgressions, operate on the same principle: signaling that authenticity has become merely another algorithmic optimization variable in the feedback loop. What varies across national contexts is not the underlying mechanism but the particular vector configurations each leader must navigate.#propaganda, #leadership, #tech-fascism
2.) The dissolution of the relationship to truth
Vectofascism does not simply produce a lie. What characterizes it is rather the production of a calculated undecidability, of a gray zone where the very status of the statement becomes indeterminable. Post-truth is not the absence of truth but its submersion in a flow of contradictory information whose sorting would require a cognitive effort exceeding available attentional capacities. This strategy exploits a fundamental asymmetry: it is always more costly in terms of cognitive resources to verify a statement than to produce it. Producing a complex lie costs a few seconds ; demystifying it can require hours of research. Vectofascism pushes this logic to the point of transforming veracity itself into a variable of algorithmic optimization. The question is no longer ‘is it true?’ but ‘what degree of veracity will maximize engagement for this specific segment?’#tech-fascism, #ground truth, #economics
3.) The designation of algorithmic scapegoats
Where historical fascism designated universal enemies of the nation, vectofascism calculates personalized enemies for each node of the network. It is customized hatred, algorithmically optimized to maximize the affective engagement of each population segment. The system does not propose a single scapegoat but an entire ecology of potential scapegoats, adapted to the preexisting affective dispositions of each user. Gamergate (2014) exemplifies this ecology of algorithmic scapegoating. Coordinated harassment campaigns emerged across Reddit, 4chan, and Twitter, targeting female game developers and critics through platform-specific vectors. Each subcommunity received algorithmically tailored grievances: misogyny framed as ‘ethics in gaming journalism’ for certain forums, conspiracy narratives for others. Discord servers and bot networks automated coordinated attacks across geographic and ideological segments. No centralized leader orchestrated this; instead, algorithmic recommendation systems and affinity clustering amplified fragmented groups into coordinated harassment networks. Gamergate demonstrated how vectofascist mechanisms function without traditional fascist organization: vectorial alignment replaces hierarchical command, algorithmic targeting replaces mass propaganda, distributed harassment replaces centralized violence.#violence, #algorithm, #decentralization
The Industrialization of the Differend
Vectofascism operates through the industrialization of the differend: a mechanism by which algorithms systematically produce situations where parties in conflict can no longer agree on the very terms of debate. Vectofascism does not need to censor the opposition; it suffices to ensure that discursive universes are sufficiently distinct that even the identification of a common opposition becomes impossible. Take two users of X/Twitter. One sees in their feed: “Scientists alarmed by warming,” “Renewable energies on the rise,” “Young climate activists”. The other sees: “Record cold in the Midwest,” “Wind turbines kill birds,” “Greta Thunberg criticizes capitalism”. Same event (COP28), two impermeable informational realities. This divergence is not accidental. Recommendation algorithms have identified each user’s initial positions and reinforce these positions through selective exposure. More troubling: the two users soon no longer share the same factual criteria. For one, the IPCC is authoritative; for the other, it is a corrupt institution. The differend therefore no longer concerns the interpretation of common facts, but the very nature of what counts as fact.#algorithm, #homophily, #datasubject
Algorithms do not merely filter information; they create epistemic bubbles, closed informational environments that generate their own internal coherence. Each community develops not only its own opinions, but its own criteria of truth, its own factual evidences, its own modes of reasoning. The QAnon movement perfectly illustrates this dynamic: it is not a simple conspiracy theory, but a complete informational ecosystem, with its sources, its investigation methods, its validation protocols. Participants do not inhabit our world: they inhabit an alternative world endowed with its own internal rationality, produced by the algorithmic aggregation of content selected to maximize emotional engagement. This mechanic operates through a simple economic imperative: maximizing emotional engagement directly correlates with advertising revenue. Platforms optimize for content that triggers intense affective responses, outrage, fear, tribal identification, because such engagement generates more data points, more attention time, more sellable impressions for advertisers. Emotional intensity becomes the key metric for surplus value extraction. This is not merely incidental but structural: the entire architecture of engagement algorithms is designed to convert affect into advertisement exposure. Vectofascist messaging succeeds precisely because it weaponizes this same mechanism, maximizing emotional charge to produce value for attention merchants.#algorithm, #attention capture, #attention rent, #datasubject
Vectofascist power is characterized by its spectrality: its capacity to deny its own existence while exercising its effects. “It’s not fascism,” is repeated, while implementing its fundamental mechanisms under different names. This denial is part of its operational power. The algorithms that constitute the infrastructure of vectofascism are literally specters: invisible to the users they modulate, present only through their effects, they haunt digital space. This spectrality partially explains the inadequacy of traditional modes of resistance. How to oppose what denies its own existence? How to resist a form of domination that presents itself not as imposition but as personalized suggestion? This invisibility is not accidental but structural: it follows directly from the very nature of high-dimensional vector spaces of ‘neural’ networks the constitute ‘AI’. How to criticize what we cannot represent? How to resist what operates in dimensions we cannot directly perceive? The power exercised in this latent space thus partially escapes our very capacity to conceptualize it.#tech-fascism, #algorithm
Alignment as Technology of Government
In the field of AI, we speak of alignment when a system succeeds in bringing its responses closer to what its designers expect from it. The most common techniques include Reinforcement Learning from Human Feedback (RLHF), where human evaluators rate the model’s responses to train it to produce content deemed appropriate.#reinforcement learning
The logic of cultural alignment is not limited to the United States; it has crystallized in similar ways across Europe, revealing an international convergence of authoritarian techniques. In America, Trump has systematized this alignment. The executive order of April 3, 2025, “Restoring Truth and Reason to American History,” justifies direct political intervention. The letter of August 12, 2025, to the Smithsonian Secretary makes the operation explicit: “This initiative aims to ensure alignment with the presidential directive to celebrate American exceptionalism, suppress divisive narratives.” It details a rigorous timeline: museums must “begin implementing necessary content corrections, replacing divisive or ideologically oriented language with unifying descriptions.” This replacement logic reproduces exactly the techniques of Reinforcement Learning from Human Feedback (RLHF). Trump extended this method to universities through budget sanctions. Harvard saw $2.2 billion frozen in April 2025, creating a negative learning signal that forced other institutions to conform. This is the structure of an algorithmic loss function: deviations generate ‘costs’ orienting optimization toward compliance. Viktor Orbán deployed identical strategies, adapted to the Hungarian context. The closure of the University of Theater and Film (2019-2021) did not result from explicit censorship, but from progressive legal and budgetary restrictions. Attacks on Central European University proceeded through legal vectors (residency requirements, funding restrictions) rather than direct repression. #reinforcement learning, #feedback, #ground truth, #cultural hegemony
This strategy functions like algorithmic deplatforming: channels are not eliminated but vectorially constrained, creating graduated exclusions. Unlike the hierarchical repression of historical fascism, vectorial alignment operates through modulation: institutions are not closed but reconfigured through calculated coercion. Culture gradually aligns toward the objectives of power, producing the appearance of freedom while neutralizing critical substance. Both regimes share a quasi-identical understanding of the cultural ‘objective function.’ For Trump: celebrating American exceptionalism; for Orbán: defending Hungarian culture against foreign influences. This formulation legitimizes intervention as protection of ‘authentic’ values. These alignments produce identical effects: concentration of cultural resources, elimination of critical positions, regression toward ideological mean. Trump rejected candidates ‘too woke’ for the Kennedy Center Honors, favoring ‘safe’ figures like Sylvester Stallone. Remarkably, these two regimes, ideologically distinct, structurally employ the same techniques: iterative schedule of evaluation-sanction-modification, measurable criteria for compliance, automated punitive signals. They transform cultural institutions into extensions of the state apparatus.#reinforcement learning, #feedback, #cultural hegemony, #woke art
This convergence reveals a profound mutation: alignment operates in invisible multidimensional latent spaces, which explains its efficacy and its capacity to reproduce internationally. The “celebration of American exceptionalism” prescribed to museums functions as an attractor optimized to capture and retain visitors’ attention. The requested “unifying descriptions” do not aim at historical accuracy, but at the production of positive affects that reinforce adherence to the national narrative. Museum curators become “engagement managers” charged with maximizing emotional attraction metrics. This subordination of culture to the attention economy transforms institutions into optimization devices. Aligned museums become ‘safe spaces’ where visitors can consume gratifying narratives without risking the destabilization of their prior certainties. This emotional securitization constitutes a form of control more subtle and more effective than direct censorship.#nationalism, #attention capture, #optimization
Alignment also operates through massive production of counterfactuals, creating alternative worlds that compete with the factual world in the attention economy. The Trumpian executive order perfectly illustrates this dynamic: it does not merely deny certain historical facts, it actively produces alternative versions of American history that become competing ‘narrative possibles.’ The directive to “replace divisive language with unifying descriptions” expresses this strategy: it is not so much about denying slavery as producing counterfactual narratives where slavery becomes a detail in a larger epic of “American exceptionalism.” These counterfactuals acquire their own consistency, a statistical reality that competes with historical reality. Visitors to aligned museums are no longer confronted with a single imposed version of history, but with a set of equiprobable narratives among which the factual version loses its specificity and critical force. This disfactuality—alteration of the perception of reality by generative technologies that blur the distinction between fact and fiction—constitutes the most sophisticated weapon of contemporary alignment.#discourse, #cultural hegemony, #history
This alignment logic extends far beyond American borders. From Budapest to Beijing, from Rome to Moscow, via Berlin, the same logic of cultural optimization unfolds according to locally adapted but structurally similar modalities. In Hungary, Viktor Orbán has developed a prototype of this governmentality through alignment. The forced closure of Central European University in 2019, compelled to exile itself to Vienna, perfectly symbolizes this reterritorialization. More significant still, the Hungarian Academy of Arts explicitly rehabilitates the figure of the socialist state artist, distributing “generous monthly allocations” to its members to “counter liberal trends in contemporary fine arts.” In Italy, Giorgia Meloni has explicated her strategy: “The left party is not the only one to have a culture.” Her Minister of Culture has systematically replaced foreign museum directors with Italians, instituting a linguistic criterion (B2 level of Italian) functioning as a barrier to internationalization. In Russia, the decree of November 2022 on “strengthening traditional Russian values” grants state representatives means to block ‘Western influences.’ In China, Xi Jinping has erected cultural alignment into a doctrine of “cultural sovereignty.” In Germany, 66 cultural events were canceled between October and December 2023, illustrating how each national territory develops its own alignment criteria according to its specific historical traumas.#liberalism, #cultural hegemony, #nationalism
The logic of closure extends to physical borders, transformed into vectorial filters. The growing practice of systematic searching of phones and computers at American borders (more than 50,000 devices inspected in 2024) is inscribed in this logic of alignment as control of cultural flows. These devices are scrutinized as latent vectors carrying non-aligned content: digital libraries, alternative social networks, critical archives. By scrutinizing browsing histories, installed applications, multimedia files, customs officers act as binary classifiers distinguishing what can ‘disturb’ the national latent space from what reinforces its alignment. The border is no longer a geographical line, but a computational threshold protecting the integrity of the national vector space. The seizure in 2024 of journalists’ phones covering pro-Palestinian demonstrations illustrates this will to maintain the American latent space sheltered from narrative ‘disturbances.’#cultural hegemony, #surveillance
The (De)generate Image and the Realism of Possibles
The emergence of diffusion technologies reveals a major ontological rupture in the history of images. Diffusion models work by removing in an iterative process pices from a noise pattern until a pixel combination emerges that to the human eye is an figurative image. While in earlier pre-trained models sample images have been used to train on specific objects, such as a house, a staircase, or a cab, in diffusion models these training samples are used to train the model by slowly adding noise. Imprinting the way that the noise is distributed in each step into the model can be later used to approximate how to remove it. Degeneration, as we understand it here, does not refer to corruption but designates the process by which images acquire an unprecedented capacity for perpetual differentiation, escaping traditional logics of the original and the copy. Diffusion models operate a revolution in our understanding of noise. In these systems, noise ceases to be the enemy of information to paradoxically become its generator. This approach reveals a fundamental property of noise in the computational context: its informational reversibility. Unlike thermodynamic entropy, which grows irreversibly, informational entropy can be reversed provided one has the appropriate decoder. We no longer merely multiply images quantitatively: we make them degenerate qualitatively by putting them in feedback. Through this process, we systematically reintroduce the original noise into the system, affirming a form of nihilism—the impossibility of obtaining a stable foundation, a pure origin, a definitive truth. The space of possibles progressively becomes the new space of images. This space is no longer constituted of discrete media but of statistics, correlations, multidimensional mathematical patterns. This new ontology can be characterized as a realism of possibles. It prevents the constitution of a unique truth by authoritative discourse and leads to a multiplication of possible regimes of truths.#Stable Diffusion, #noise, #statistical model, #generative
Elon Musk’s salute during Trump’s inauguration in January 2025 perfectly illustrates this transformation. To normalize this gesture, images appeared on social networks representing various historical figures, including Martin Luther King, with arms extended. It is by operating a decontextualization and comparison with other documents that the specificity of the gesture disappeared according to a procedure analogous to what happens in the latent space of an AI. One could provide an ‘AI’ with thousands of images of Nazi salutes and, by comparing them statistically, it would transform these discrete documents into continuous statistical vectors allowing the generation of what, at minimum and maximum, such a salute is for us. It doesn’t matter the context, only the automation of resemblance counts. Finally, the software would be capable of generating an image attributing a Nazi salute to anyone, thereby neutralizing the significant scope of such a gesture. We have moved from the image as copy within the framework of technical reproducibility inherited from the industrial revolution to the image as production of a statistical flow, of which it is only one possible iteration. Images are no longer proofs, but the hallucinations of a statistical model to which we have no access. The hallucinatory consequence is that Musk’s gesture has become normal, insignificant, a sign among other vectors in the immensity of the possibilities of the latent space that has become our reality.#ground truth, #generative
The meme, this singular form of collective creativity that has structured the internet for two decades, operates according to a logic that Derrida would have recognized: that of différance, this simultaneous movement of differentiation and temporization. Each meme differs (distinguishes itself) while deferring (postponing) meaning in a chain of infinite reappropriations. Take the Drake meme. The original image of the rapper—two close-ups expressing disapproval then approval—becomes a formal matrix that millions of users reinvest to express their cultural preferences. Each appropriation reveals new potentialities of the format. When users replace Drake with Chopin rejecting pop music in favor of classical, or with a cat preferring kibble to vegetables, they don’t just change the referent, but explore the expressive limits of binary contrast itself. This economy of différance supposes a specific temporal regime: that of deferred propagation, of collective maturation, of semantic accumulation. A traditional meme takes days, sometimes weeks to propagate and mutate through communities. The time of the meme is that of productive anachronism. It allows heterogeneous elements to meet, to contaminate each other mutually, to produce the new through collision of distinct temporal universes.#meme, #ground truth
Generative AI radically transforms this visual economy. In the latent space of a generative model, there are no longer discrete images. A ‘white Persian cat’ could be represented by a vector over thousands of dimensions. A ‘Bengal tiger’ would be a few coordinates away. Between these points exists an infinite continuum of intermediate positions. By interpolating between the cat and the tiger, one obtains hybrid creatures that never existed, but remain visually coherent. The distinction between meme generation and distribution proves essential to understanding this rupture. Traditional memes operated through an economy of différance: discrete images propagated with temporal deferral, mutating through successive détournement. What characterized meme circulation was distribution—collective propagation through networks, progressive semantic maturation across communities. GenAI inverts this logic entirely. Models trained on massive datasets generate instantaneously thousands of variations. Yet this generative speed conceals a crucial material reality: it requires colossal computational infrastructure—server farms, GPUs, energy systems—that compresses temporal advantage into infrastructural dominance. This represents a new form of primitive accumulation: the non-consented appropriation of all human cultural production circulating online, transformed into statistical patterns.#generative, #meme, #variations, #infrastructure
GenAI thus abolishes the meme’s deferred temporality, replacing collective maturation with instantaneous probabilistic calculation. What appears as technical acceleration masks an infrastructure of cultural seizure—vectorial extraction of the entire visual commons into proprietary latent spaces. LAION-5B, one of the main datasets used to train the generative AI model Stable Diffusion, contains 5.85 billion image-text pairs extracted from the web. Each image uploaded to Instagram, each meme shared on Reddit potentially becomes a micro-contribution to the cognitive capital of these systems. It is the primitive accumulation of meaning: the massive and often non-consensual appropriation of human cultural production. Couldry and Mejias term contemporary data extraction ‘digital colonialism,’ drawing parallels to historical primitive accumulation. Just as European colonizers seized land from indigenous populations, tech corporations systematically extract value from social activity without compensation or collective ownership. The Amazon seizure of Native lands, the enclosure movements of early capitalism, and contemporary data harvesting follow identical logics: appropriation of commons through legal-technical mechanisms. Digital colonialism operates through terms-of-service agreements and algorithmic opacity rather than explicit violence, yet achieves equivalent dispossession. Recognizing this genealogy reveals vectofascism not as aberration but as intensification of capital’s foundational appropriative mechanisms.#LAION 5-B, #Stable Diffusion, #primitive accumulation, #digital colonialism, #Nick Couldry, #Ulises Ali Mejias
Generative AI inaugurates a post-memetic regime. It reconfigures the structural conditions that made memetic culture possible: the discreteness of images, the deferred temporality, the significant temporal short-circuits. But perhaps it is appropriate to think of this transformation not as an external threat to the human, but as the culmination of a process of reciprocal alienation already at work. Generative AI does not alienate us so much as it reveals and accentuates a constitutive alienation: we have always already alienated ourselves in our technical productions, just as they have alienated themselves in us. This reciprocal alienation should not be understood in the classical sense of pure dispossession, but as a process of mutual co-constitution where the human and the technical transform each other without either remaining identical to itself. Generative AI returns to us an image of ourselves that is neither entirely us nor entirely other. In its outputs, we recognize our cultural patterns digested by probabilistic logic, transformed into something that resembles us while being foreign to us. It is the vertigo of technical uncanny strangeness.#generative, #alienation, #patterns
Vectofascist Aesthetics
The emergence of a new fascist aesthetics in the current technological context differs radically from historical manifestations. If totalitarian regimes of the twentieth century had privileged a monumental classicism, contemporary fascist tendencies adopt an aesthetic orientation characterized by the grotesque, kitsch, and a particular appropriation of pop culture. Generative artificial intelligence appears as particularly fertile ground for the development of this sensibility. Neural networks, trained on immense corpora of existing images, naturally produce referential results. This technical characteristic corresponds precisely to a visual culture that developed on certain platforms like 4chan, characterized by appropriation, diverted citation, and escalation. The NFT-artist Beeple’s work perfectly illustrates this tendency. His production, characterized by saturated, grotesque, and deliberately excessive digital images, could be interpreted as a critical approach. However, this reading neglects the fundamental dynamic: the systematic leveling of all references, the reduction of all positions to a generalized equivalence. By turning everything into derision according to often vulgar modalities, this approach does not propose a critique but institutes an integral relativism where value distinctions become impossible. Crude humor and visual escalation do not function as tools for critical distancing but as instruments for deactivating judgment.#aesthetics, #kitsch, #Beeple
Faced with this situation, two distinct orientations emerge. The first path extends the legacy of pop art by adapting it to contemporary digital cultural codes. This approach finds in AI an instrument to multiply references, latent spaces naturally facilitating this referential accumulation. This tendency can be qualified as ‘default aesthetics.’ This orientation constitutes a ‘factual fascism’ insofar as these images enclose us in a referential circularity, a nauseating repetition of the already-known. They open no alternative horizon but confirm and reinforce existing structures of domination by making them omnipresent and inevitable. The second path proposes a more radical exploration of the potentialities of latent spaces, seeking to bring forth visual configurations that escape established references. This orientation can be qualified as ‘factitious aesthetics,’ not in the pejorative sense of concealed falsity, but in the sense of an assumed counterfactuality, of an exploration of possibles that does not claim to reproduce the real but to exceed it. By deliberately integrating a part of noise in diffusion processes or by exploiting bugs and emergent behaviors of systems, this approach departs from neofascist kitsch aesthetics to venture toward unexplored visual territories. The simulacrum is not presented there as a truth but affirms itself as simulacrum, thus instituting a reflexive relationship that escapes the referential circularity of factual fascism.#aesthetics, #tech-fascism, #ground truth, #Stable Diffusion
The example of the video on Gaza, initially conceived as a ‘critical joke’ then reposted and diverted by Donald Trump, perfectly illustrates the dynamic of reversibility that characterizes this aesthetic regime. In this environment, no critical intention remains stable; all content can be reappropriated, diverted, inverted. Contents become ‘bifid’ or ‘chiral,’ carrying a constitutive ambivalence that makes them appropriable by diametrically opposed political positions. This generalized reversibility opens particularly problematic political perspectives. By establishing equivalences between profoundly distinct historical phenomena, this relativistic logic allows the gradual rehabilitation of political positions previously considered unacceptable.#relativism, #backlash
Part II: Vectopolitics
https://carrier-bag.net/vectofascism-part-2-vectopolitics/
