With the diffusion of Sora and Veo, publicly-accessible AI is now producing accurate representations of “real” video with ease. Though these images are not yet indistinguishable from reality, this is certainly temporary. What unsettles me most is not the deepfaking of reality but something more recursive: the “creative destruction” of creativity. Previous waves of economic change have destroyed industries through the birth of new ones. But we’re now automating the capacity to “create” anew, ceding our role as authors of human cultural evolution.
It’s widely believed there is something profoundly unique about human creativity. This is a comforting tale. Yet, humans already cannot distinguish between poetry produced by human beings and LLMs; productions like The Velvet Sundown, an AI-generated band, are already proof-of-concept. The uncomfortable truth? This is the worst these systems will ever be.
Throughout many philosophical traditions, intelligence has been treated as humanity’s privileged distinction from ‘lesser’ organisms. We grieve a dying dog, but crush insects without much awareness. Few of us mourn when a computer “dies”. Our empathy follows a hierarchy built principally on degrees of “consciousness”, on the assumption that the richness of our “inner lives” are what separates the master from his dog.
Searle’s Chinese Room once guarded this hierarchy: A machine can output perfect Chinese by blindly following a set of rules; Searle shows he could do the same by hand while understanding none of it. Machines could imitate meaning, but not possess it; syntax was not semantics. This distinction matters theoretically, but it dissolves under market forces. What happens when non-human outputs are indistinguishable from the “real thing” and, crucially, cheaper? Firms won’t care for ‘real’ understanding. If the output suffices, philosophical distinctions become economically irrelevant, however much one may believe it matters in the abstract.
Our preconceived hierarchy is collapsing – not because machines have grown to meet us, but because the systems humans have built do not care for “consciousness”. Man, machine and animal collapse into mere inputs. Whatever mysterious uniqueness the human carries, it it not profitable.
Technologists have long been promising an affluent society free from the binds of monotonous work, one where we are free to pursue art and passion projects—Keynes predicted in 1930, thanks to machines, we’d have a 15-hour work week within a century. Bill Gates recently said AI will deliver a 2-day work week within a decade. No such post-work predictions have come to pass. Waves of technological innovation have promised liberation, but our structures of work have remained remarkably stable.
Why might AI and automation be different?
The answer: complementary vs. substitutive automation. The GPS didn’t replace drivers, it made them more efficient. Autonomous vehicles seek to eliminate drivers as an occupation. Previous waves of automation were largely complementary. Even when jobs disappeared, human cognitive labour allowed for workers to retrain or adapt to new functions.
New forms of displacement are increasingly categorical, not occupational. Instead of tasks, we may begin automating cognitive labour itself – the very skill that enables humans to adapt to new roles. If we automate adaptation, “reskilling” becomes incoherent.
This is why standard technophile narratives ring hollow. If technology brings “liberation”, it will be primarily for capital ‒ unbound from its relation to (human) labour.
Humanists argue that some things remain valuable because humans do them: a jazz band, a haircut, a game of chess played by Magnus Carlsen rather than a perfect engine. These rely on what Polanyi called ‘tacit knowledge’, the skills we “know more than we can tell”: intuition, judgment, embodied sense.
But tacit knowledge may still be economically doomed. A machine doesn’t need to feel an egg crack to replicate the motion at scale. Human uniqueness can persist philosophically while dying economically.
This is what humanists fail to comprehend. Creativity’s exceptionality won’t shield it, because markets prize exchange-value, not essence. If imitation is “good enough,” interiority ceases to matter. The true conflict is not human vs. machine, but human vs. an economic system that sees both as interchangeable.
AI is not an aberration within capitalism, but its logical conclusion. A system built on the unrelenting minimisation of labour costs would struggle to dream up anything more true to itself than a machine which imitates human output without wages, fatigue, or complaint. There is an exquisite irony: these systems are made to replace workers by first being trained on those same workers’ uncompensated labour, “scraping” centuries of demanding human creative labour to shape the very apparatus that will render would-be creators economically redundant. Marx predicted this struggle for control over the “power of knowledge”; he foresaw this theft of the “general intellect” – our collective social commons – enclosed like land was in primitive times.
You can see this logic playing out with the ongoing lawsuit against OpenAI; Anthropic has already been forced to cede to authors. Several ongoing legal challenges allege that these models are built on unlicensed and uncompensated labour.
The potential “great novelist” of the future won’t be defeated by lack of talent or obscurity, they will instead be smothered by the very brilliance of their forebears that was meant to inspire them. Masterpieces become mulch for machines that never sleep nor think, past talent is to be exploited by the now-talentless, chopped up and sold back to us as “new”. Zola, Orwell, Austen, Freud and Foster Wallace ‒ all become the unwitting and unconsenting gravediggers of their own successors.
Literature is not alone. Music will be mined for motifs, movies for mise-en-scene, painting for its pigments: each form turned into data for a system capable of imitating everything except the necessity that once compelled creation.
Some assure us this is alarmism, that “true art” will weather this storm. Opponents believe there will always be a market for ‘authenticity’.. But consider what consumers already tolerate: auto-tuned singing, ghost-written memoirs, algorithmically-curated playlists. The question is not whether anyone will value human art, but whether enough people will to sustain an economic infrastructure where creative work remains a viable source of employment for millions rather than merely a hobby for the decommodified wealthy.
Our “culture” has become so hungry for frictionless production that it is consuming the very conditions that first birthed its renewal. The breakthroughs of yesterday are repurposed as tomorrow’s barricades. Arendt warned in her essay The Crisis in Culture, consumer society’s “central attitude toward all objects… spells ruin to everything it touches.”
And here is the even deeper contradiction: AI’s appropriation is self-cannibalising. These models require human creativity as “training data”. Eventually, they will require new art if they are to remain “current.” Yet, by wrecking the economic basis for creative labour, they may come to destroy the conditions for their own renewal. You cannot indefinitely obtain clean water from a well you’ve poisoned. Of all people, the Pope put it succinctly, “algorithms… repeat what ‘works,’ but art opens up what is possible”.
Perhaps we’re asking the wrong question about what makes us human. There’s a scene I can’t shake from Werner Herzog’s documentary Encounters at the End of the World: the ‘deranged penguin’. A solitary bird, for no apparent reason, turns from their group and marches toward an inland mountain range. Even if humans return the penguin, the same thing reoccurs – it marches inevitably towards certain death.
Even in the cold calculus of nature, the bird is not pursuing survival nor meaning. In its pointless pilgrimage, I recognise something familiar. That penguin exposes a truth we seldom admit: the irrational, futile, self-destructive impulse is not a human monopoly. Nature too breeds its own madmen.
We too are strolling leisurely toward our own obsolescence; this is not for lack of intelligence, creativity or consciousness, but because the systems satiating our desire have their own logic, indifferent to where they take us. We discover ourselves in the penguin’s march – not human exceptionalism, but human deficiency. We attach meaning to its doomed advance because mythologising entropy is our reflex.
We watch that penguin and we narrate it. We project concepts onto it, call it deranged or inexplicable or heartbreaking. We impose a story where there may only be chemistry. For me, that is “the human” if anything is; not intelligence or creativity, but the way we interpret change. We make myths out of what we feel powerless to change; we narrate our own growth and decay to make it manageable.
For centuries, our narrative impulse was valuable precisely because it was both scarce and labour-intensive. A symphony required a composer; a novel demanded an author. In Marx’s terms, the value of any commodity is based on the socially necessary labour time (SNLT) ‒ the average labour-time society needs to produce a given commodity with prevailing levels of technique.
Markets rewarded storytellers not because humanity was essentially creative, but because creativity was essentially human.
Markets rewarded storytellers not because humanity was essentially creative, but because creativity was essentially human. AI won’t decommodify meaning-making, it’ll simply end its scarcity, collapsing its SNLT. Under capitalism, this is equivalent to collapsing a good’s economic value, and thus profit derived from it. Stories will no longer need (new) storytellers. The essence of creativity is now unmoored from humanity.
None of this demands revolution nor ignores the merits of capitalism or AI adoption. The tragedy is that everything is likely to continue exactly as it is. Society will luxuriate in the ease of new technology, normalising each incursion as it did with social media, smartphones, and the entire surveillance architecture we no longer resist. The question of “human value” will turn quaint, sentimental, economically inappropriate; “manpower” may become akin to the term “horsepower”.
Fisher was right: it is easier to imagine the end of the world than the end of capitalism. We already envision a future of AI-generated culture in which humans become passive consumers. Yet we fail to imagine collective ownership of these systems, compensation for our “general intellect”, or democratic control over how our creative labour, our very society’s history, is scraped, privatised, and rented back to us.
So, we can continue accepting that value is largely what the market says it is; that the interior life—memory, longing, pain, foolishness, all such textures that make us feel human—belongs to that bizarre category of priceless things which, deficient of price, lack shelter in a world where only price confers value. It’ll live alongside care that yields no return, time that produces nothing of exchange-value, and irrational rituals sustained because the abandonment of them would make us unrecognisable to ourselves.
We preserve these things right up until convenience demands otherwise. Here we discover what our economic behaviours have acknowledged long ago: most of us were never that serious about defending what we claim makes us human. Perhaps that’s the most human deficiency of them all.
The question we must ask ourselves is not “What makes us human?”
It is instead:
Are we prepared to live in a world where the answer doesn’t matter?
Leave a comment