At the start of August, the Tricontinental Institute for Social Research unveiled a new and unconventional weapon in the class war: a bespoke Marxist Artificial Intelligence dubbed RedMarx. To motivate its usage, the team behind RedMarx posited “Mariana”, a hypothetical Brazilian political economist tasked by an agricultural workers’ union to delineate “the relation between the socialisation of labour and the rising organic composition of capital”. While conducting her research, Mariana naturally consults ChatGPT and Gemini – but the answers she is given are inevitably inadequate. The Tricontinental Institute’s ambition is to develop an AI that will answer adequately.
At the core of RedMarx is a hierarchical network of interconnected concepts; what its makers term a “layered conceptual knowledge graph of Marxist political economy”. When a question is asked, the model first creates a sub-graph of pertinent concepts. Then it searches through a corpus of Marxist texts and parses the resultant data through the sub-graph’s conceptual framework. The result is, supposedly, a Marxist interpretation of Marxism.
I don’t know if RedMarx really ‘works’ or not. The tool is still in the training stage, though the creators have provided some worked examples of the output on their GitHub repository: “The tension between formal subsumption (capital commands but does not transform the labor process) and real subsumption (capital reshapes the labor process in its own image) is resolved dynamically: the social productive forces generated under formal subsumption create the material basis and the competitive pressure for the transition to real subsumption.” One might question how useful the agricultural workers’ union would find this answer.
Whatever hypothetical Brazilian labourers might think, the actual reaction of the online Left has been harsh. At the time of writing, 12 of the 18 top-level comments on the Institute’s Instagram post were clearly negative. Examples include: “This is the opposite of class solidarity”, “generative AI cannot analyze or interpret!”, and “new and exciting ways to distort marx”. Responses to the X post have been similarly critical in tone.
Comment sections are not necessarily representative of course, and they have a marked tendency towards the gratuitously brutal. But, as far as I can tell, the more considered responses on the Left have been generally negative as well. On the Red Lens Substack, Serge Jordan writes that the RedMarx project “rests on a fundamental misunderstanding of what Marxism is”. Marxism, Jordan contends, is a method of revolutionary action, not “a catalogue of texts and concepts”. Any such corpus-based tool will inevitably exclude the “lived experience of the working class” – the lifeblood of Marxism. Artificial Intelligence is doomed to be a Frankenstein’s monster, a death-like mockery of the living movement it claims to serve.
Wrecking the machine
All new technology leads to backlash, and yet not all backlash is driven by pearl-clutching or fear of the new. Luddism, as Eric Hobsbawm pointed out, was far from a mindless effort to reverse the tide of technological change. Machine-wrecking was a deliberate and frequently successful strategy employed by workers in the inaugural phases of the Industrial Revolution.
Should we be smashing data servers just as our 19th-century forebears smashed power looms? There are indeed good reasons to distrust AI. In the early days of its mass adoption, the most typical criticism was that it was a ‘stochastic parrot’: a device for stringing together words on the basis of probabilistic connections that would never provide more than a thin and unreliable simulacrum of human intelligence. We laughed at its bloopers and cringed at its grotesque six-fingered hallucinations.
But no one doubts the power of Artificial Intelligence anymore. AI is solving decades-old maths problems; it’s winning prestigious literary awards; it’s producing poetry that is rated more beautiful and more authentically human than actual human productions. It’s stealing our jobs and taking our women. It has a greater than 10% probability of destroying humanity. Suddenly it doesn’t seem so funny. Even if it doesn’t reach the astounding heights that its most starry-eyed proponents claim for it, Artificial Intelligence is clearly a powerful new technology that will fundamentally transform many aspects of our world.
Unfortunately, this new technology is concentrated in the hands of malevolent billionaires whose visions of world domination bear a striking similarity to the musings of someone halfway through lowering James Bond into a piranha pit. Men such as Peter Thiel, Elon Musk, and Palantir boss Alex Karp, whose manifesto proclaims that “one age of deterrence, the atomic age, is ending, and a new era of deterrence built on AI is set to begin”.
While other ‘founders’, such as Dario Amodei, are more careful in their public presentation, their commitment to Western imperialism is clear. Just last year, senior executives at Meta, Palantir, and OpenAI were sworn in as members of a new reserve unit of the US Army whose aim is to deepen integration between big tech and the military. Western AI tech has already been implicated in atrocities in Palestine, and the murder of over 100 schoolgirls at the start of the US-Israeli war on Iran.
Downfall of the clerisy
But the Left’s pessimism towards Artificial Intelligence seems to run deeper than just a critique of how it’s being used or who owns it.
In Jacobin, Sam Kriss warns that AI will leave us “in the ruins of a sophisticated literary culture that [has], for the most part, vanished from the world”. As we depend more on the brains in our devices, we will lose not just fluency of reading but also the ability to understand metaphors or to reason through abstract scenarios. With the population at large hooked up to “an AI-powered pleasure machine that will keep them in a state of permanent hedonic bliss”, the way will be clear for Silicon Valley to exclude the masses from politics, ushering in a new age of untrammelled oligarchy.
The editors of n+1, a US-based leftist literary magazine, are equally pessimistic. In an article subtitled “It’s OK to be a Luddite!”, they warn that AI poses a threat to the livelihoods of intellectuals and other members of the “professionally and semiprofessionally literary”. Like Kriss, they discuss the proliferation of AI use among college undergraduates and the consequent despair of humanities academics, who are supposed to be nourishing the soft yet crucial skills of clear communication and critical analysis.
Nonetheless, the editors argue, resignation is premature. AI will never replace the “charm, surprise, and strangeness” of real human writing. Rather than surrendering, the editors call for a refusal to write or publish “AI bullshit”; even to admit to the temptation of using it should be verboten. AI should be banished from the classroom, and colleagues who use it should be shamed. “Put plainly,” they say, “you should feel bad for using AI”.
It is not hard to see merit in these arguments, for the development of Artificial Intelligence seems to have pushed the frontier of capitalist alienation beyond the physical limits of body and time. This incursion into the human constitution touches directly on Karl Marx’s initial critique of the capitalist mode of production – a critique which was not centred on the role of economic exploitation per se, but on capitalism’s tendency to “estrange” or alienate those subjects sucked into its ever-expanding vortex.
Humans, Marx thought, have an innate tendency to use their capacities for the purpose of self-directed production. Capitalism robs us of the right to direct those capacities in accordance with our own flourishing, and thereby reorientates our life’s work towards the goal of capital accumulation, leaving us only the solace of consuming others’ alienated labour in the mystified and ephemeral form of the consumer commodity. Marx movingly describes the deepening dynamic of alienation this process engenders: “The more the worker spends himself, the more powerful becomes the alien world of objects which he creates over and against himself, the poorer he himself – his inner world – becomes, the less belongs to him as his own.”
Yet until now there has always existed a frontier beyond which the alienating forces of capital could never penetrate: our internal lives; that is, our capacity for imagination and reason. From this kernel of humanity – call it soul, heart, or species-being – springs the basis of every act of resistance, and every emancipatory conception. With AI, capitalism can cross this frontier and, like hapless Wallachian peasants, we have willingly invited the vampire across the threshold of our home. Write my essay. Plan my life. Choose a birthday present for my mother. Tell me why my husband cheated on me. Even if we don’t want to use it, we know that others will. And if we still refuse to outsource our intelligence, we will be left behind, secular Amish in the new digital world.
An inexplicable emptiness
But this is not the first time that capital’s greasy fingers have fumbled at our shirt buttons. In his 1935 essay The Work of Art in the Age of Mechanical Reproduction, Walter Benjamin predicted that the ability to record sound and motion entailed a “far-reaching liquidation” of the best of humanity’s cultural heritage: Shakespeare, Rembrandt, Beethoven.
Of these new technologies, the cinematic form was the most destructive, for it severed the connection between actor and audience. Through the process of recording and editing, the actor’s movements would be cauterised into snippets of human action, sterile representations that could be stored and assembled as if on a factory construction line. The result is that the actor feels (here Benjamin quotes the Italian dramatist Luigi Pirandello) an “inexplicable emptiness”.
Because a film negative could produce an indefinite number of copies, it no longer made sense to speak of an ‘authentic’ version. The negation of authenticity spelled the abolition of art’s ‘aura’ – the reverential effect produced by the sense of distance between the spectator and a work of art. From now on, the masses would consume their entertainment in a mood of distraction rather than contemplation. Rather than attend with rapt solemnity, they would notice incidentally. Both artist and audience would lose the hallowed concept of originality.
In Benjamin’s view this destruction of originality was good. Before mechanical reproduction, works of art possessed an ‘aura’ that derived ultimately from art’s origin in ritual and religion. The persistent influence of this ritualistic character gave art a regressive function in the era of class struggle. The abolition of ‘authenticity’ therefore implied the abolition of this religious orientation, clearing the way for revolutionary transformation.
Free from the illusion of a private and individual experience of aura – akin, perhaps, to prayer – the working masses would experience fully the radical implications of their mass-ness. For just when specialisation and urbanisation appeared to have locked the workers up in their own private sphere of alienation, cinema’s explosive reorganisation of experience had burst the prison gates with the “dynamite of the tenth of a second”. The distinction between actor and spectator was to be erased – just as the modern press had abolished the distinction between reader and writer. At last the individual could enjoy their “legitimate claim to being reproduced”.
A passion for mechanical improvements
There could not be a more striking contrast between Benjamin’s attitude towards the new technology of his day and how the contemporary heirs of his political tradition confront that of our own. Far from wringing his hands over the implications for our souls, Benjamin welcomed how film altered our sense of self and castigated those who expressed reservations at its impact on human consciousness. In the same essay, Benjamin lambasts the Czech writer Franz Werfel as an “ultrareactionary” due to the latter’s wish for film to replicate the “supernatural” quality of earlier artforms. One would presume that this designation implied that Werfel was a fascist, or at least a sympathiser of that movement which would hound Benjamin to a fatal morphine overdose in a Catalan hotel. In fact, Werfel was of Jewish origin, an associate of Kafka and Kraus. His books were burned by the Nazis, from whom he fled across the Atlantic. But for Benjamin, Werfel was an “ultrareactionary” – not by political conviction but owing to his denial of the emancipatory potential of the new.
Benjamin was by no means an outlier in his embrace of mechanical progress. On the contrary, his optimism towards technology and its far-reaching implications for human consciousness was very much in line with how others on the Left were thinking at the time. In those years, Marxism saw itself as aligned with Darwin’s evolutionary theory and Tesla’s electro-magnetic innovations, part of a broader intellectual movement that would break the fetters of superstition and servile tradition.
In Literature and Revolution for example, Leon Trotsky predicted that socialism would inspire a “passion for mechanical improvements”, unleashing godlike powers which would provide a new source of inspiration for artistic work and a basis for astounding feats of engineering and construction. No aspect of nature would be left unconquered, from the deepest trenches of the ocean to the highest mountain peaks. But the real work of liberation, Trotsky claimed, would be wrought on the body and inscribed on the soul. Man would become “immeasurably stronger, wiser and subtler; his body will become more harmonized, his movements more rhythmic, his voice more musical”. The ultimate goal: creation of “a higher social biologic type, or, if you please, a superman”.
Pessimism for the machine
One hundred-odd years later and you will be a long time searching for such unalloyed optimism on the Left. Draw a political quadrant so that the horizontal axis runs from left to right and the vertical descends from techno-optimism to techno-pessimism. The top quarter of the left-hand side would appear very sparsely populated indeed – even the Tricontinental Institute do not claim to be paving the way for the socialist Übermensch. What has changed?
The short answer is that the Left has been on the back foot since the end of the last century. While undoubtedly authoritarian, the Soviet Union’s existence suggested the possibility of constructing another system beyond capitalism. Enthusiasm for the socialist motherland is apparent throughout The Work of Art in the Age of Mechanical Reproduction; Benjamin assures us, somewhat naively, that the Soviets had already effectively abolished the distinction between artist and spectator.
From today’s vantage point however, capitalism seems inevitable; socialism, a brief historical aberration. The only people who dare to dream of building impossible cities are crude philistines like Elon Musk. Meanwhile, in the face of transformative technological developments, the Left has retreated into a defensive crouch, bracing for a doomed yet noble fight for the preservation of the last residue of unspoiled humanity.
But doesn’t this project of preserving a pure, hermetic individuality seem somewhat… bourgeois? And doesn’t resistance to new tools of knowledge production seem a little ‘reactionary’ – in Benjamin’s sense at least? Benjamin certainly would not be slow in identifying the religious overtones of leftist jeremiads, whose insistence that even the temptation to use AI must be stamped out is somewhat reminiscent of a spittle-flecked bible-thumper raging against the dangers of self-pollution. Go back far enough and you will encounter Socrates warning that the adoption of writing would destroy knowledge. And we all know what he thought of the masses.
The dwarf in the machine
Yet towards the end Benjamin came to reverse his optimistic view that technological and social progress would necessarily march together. In his Theses on the Philosophy of History, unpublished in his lifetime (he gave it to Hannah Arendt in Marseilles, who delivered it to Horkheimer and Adorno), Benjamin opens with a startling image: “It is well-known that an automaton once existed, which was so constructed that it could counter any move of a chess-player with a counter-move, and thereby assure itself of victory in the match.”
Concretely, this automaton was a puppet dressed in Turkish robes, with a hookah perched in its inanimate mouth, sat before a chess board over a broad table. But the automaton only appeared to be automated. Because crouched beneath the table, and hidden by a clever system of mirrors, was a “hunchbacked dwarf” who manipulated the puppet’s arms through a system of strings.
The puppet, Benjamin announced, was “historical materialism”. The dwarf was theology, “which as everyone knows is small and ugly and must be kept out of sight”. The rest of the text is likewise shot through with religious and apocalyptic terminology: Resurrection, the Last Judgment, the Kingdom of God. In 1935, religion was an out-moded fetter on human development; in 1940, it had become a last light guttering in the dark wind from Berlin.
Temporally speaking, Benjamin’s last work is determinedly orientated towards the past. Progress is a storm of catastrophes, rubble piled on rubble, watched over by the forwards-flying, backwards-facing angel of history. Perhaps, Benjamin suggests, something might be rescued from the wreckage; finer, spiritual things may glimmer between the hammer blows of contending forces.
The change in tone may be explained in part by the grim historical context in which Benjamin wrote this text. France had fallen and the author found himself fleeing from the Nazis. The Soviet Union – where the contradiction between the production and consumption of art was supposed to have been abolished – had a year prior signed a pact with the very forces who would hunt Benjamin to his death. There is not one mention of the Soviet Union in this literary last testament.
But fascism is mentioned several times. Benjamin had also mentioned it in Mechanical Reproduction. There he acknowledged that fascism had also embraced the possibilities of modern technology. But the difference between fascism and socialism, he had contended, was that the former sought to give people only the ability to express themselves in jingoistic fantasy, while denying their right to abolish the economic system which exploited them. Fascism, for Benjamin, was “the introduction of aesthetics into political life”.
In this sense at least, Silicon Valley’s oligarchs can be legitimately categorised as fascist in orientation, whatever label they consciously embrace. Social media, and now artificial intelligence, muffle the political instinct by offering zones of depoliticised expression where radical content is quietly censored. Just ask ChatGPT whether Gaza was a genocide.
A storm blowing from paradise
But regardless of our outlook, we can no more halt the rise of artificial intelligence than we could have halted the steam engine or the container ship. There is certainly room for tactical Luddism as a short-term stratagem against the predations of the Silicon elite. Palantir, for example, should be ripped from any institution into which it attempts to insert its tentacles. Still, Mariana – the hypothetical Brazilian political economist whom the RedMarx team invited us to imagine – would very likely have used AI in the course of her labours. I certainly have in mine. Here is Claude’s summary of what I asked it to produce while drafting this piece:
“You asked me to recall the “stochastic parrot” critique of AI; find and verify the Trotsky “new man” passage from Literature and Revolution; locate critiques of the RedMarx project and read the article into context; explain who Tricontinental and Global South Insights are; compile lists of leftist critiques of AI, then narrow to recent articles taking an epistemological angle; count the negative comments on an Instagram post; and find and cite the Marx “Estranged Labour” quotation from the 1844 Manuscripts”.
Did my usage of AI make this article worse? I don’t think so. The ring does not necessarily corrupt its bearer; the beggar may still string the bow. Hope for the future lies in removing oligarchic control over information technology – and every other aspect of our society. It lies in the prohibition of surveillance and the sustainable construction of server infrastructure. It lies in a balanced path of economic development – not a manic race to mass layoffs and spiralling debt. In the short-term, it lies in using the tools we have in a responsible way. Principles of conscious and critical usage should be disseminated. Systems should be designed to incentivise the use of AI as an aid for learning rather than a replacement for it. And the technology of Large Language Models should be harnessed for its emancipatory potential – just as an earlier generation of socialists dreamed of using the film camera and the steam engine to expand the frontier of human creativity. RedMarx is only a prototype. But it is a step in the right direction.
William Foley is a research fellow based at Universidad Carlos III de Madrid. He is currently researching the spread of religions, the evolution of inequality, and the role of luck in life outcomes.