Summary: Radu Jude made a nearly three-hour Dracula that both fattens and skewers AI: he calls the output “gross and slimy” while feeding his film with the very images he mocks. The result is loud, abrasive, and deliberate — full-frontal nudity, gore, broken anatomy, and AI-generated Draculas that look like nightmares from a badly trained model. Jude frames the experiment as curiosity, thrift, and provocation. He also knows it will anger people who see generative systems as a theft machine. This post lays out what he did, why he did it, what people are saying, and what the argument really is about: art, labor, memory, and who gets to make culture when algorithms drink from human work.
What Radu Jude made and why it matters
Radu Jude is not an accidental provocateur. He won the Golden Bear in 2021 for Bad Luck Banging or Loony Porn, a film that used porn, public outrage, and the empty streets of pandemic Bucharest to force viewers to face cultural hypocrisies. His next film, Do Not Expect Too Much From the End of the World, kept the same mood—post-Covid alienation, a protagonist working a dull corporate job who moonlights as a misogynist TikTok persona. Jude works on what he calls “the margins of Europe,” yet his themes travel: anxiety about constant connection, the clash between high intellectual discourse and crude humor, the mix of cinema-school theory and street-level vulgarity.
That track record matters because it frames Dracula. This is not a conventional vampire picture. It is a meta-film: a filmmaker within the film uses an AI app to generate short Dracula vignettes when he runs out of ideas. The AI images are explicit, bizarre, often grotesque—figures with three hands, heads multiplied, genitalia misplaced—kept as-is, not fixed. Jude says the images are “slimy” and “gross,” and he deliberately leaves their mistakes. He calls the result “AI poetry.”
How he used AI — and what he tried to make it do
Jude’s process mixed provocation, thrift, and experimentation. When ChatGPT and image generators picked up public attention, he started testing prompts. He even tried to push the model into producing a pornographic Dracula set in Auschwitz; the AI refused. That refusal was a trigger. Jude then pretended to be the AI, producing the material the machine would not. He also used AI imagery to cut costs: Romania’s film industry is small, budgets are tight, and AI offered a cheap visual armature.
On Zoom at a New York Film Festival screening, Jude appeared with an AI background of Donald Trump waving an AR-15 while riding a cartoon cat—an image meant to shock as much as amuse. A skeptical cinephile called him “on fraud watch.” The exchange captures the split reaction: is he a smart, playful critic or a fraud using problematic tools while mocking them?
Gross and slimy — mirrored phrase, repeated
Jude said the AI results looked “gross and slimy.” Gross and slimy. He kept those images not to hide the errors but to showcase them. Rather than polish a photorealistic fake, he left the glitches—three hands, four heads, genitalia in the chest—to make a point about what algorithmic generation produces when it’s fed human data and no moral compass. That deliberate ugliness becomes his material. He calls it “digital poetry.”
Ask yourself: what does it mean to accept the machine’s mistakes as aesthetic choices? What does it mean to keep “gross and slimy” rather than erase it? Are those errors poetic truth or cheap provocation? Those are open questions Jude wants you to sit with.
Art as critique and the paradox of using the tool you mock
The central tension is straightforward: can you use a technology to criticize it without betraying your critique? No, it is not a simple either-or. Jude leans into the paradox. His earlier films used staged violence or pornography to unmask repression and hypocrisy. Dracula uses AI imagery to reveal how the technology looks when it is fed culture, commerce, and human failure.
Some say any use of AI legitimizes it. Others say critique must engage the object it criticizes. Jude comes from a place where you can afford to be experimental—Romania’s industry is tiny, he says, so there’s “nothing to lose.” In the United States, the stakes are different: millions and billions of dollars, jobs, unions, and culture industries that can be gutted by cheap automation. That difference in stakes explains part of the outrage.
Economics, aesthetics, and environmental concerns
Jude acknowledges three problems people complain about: economic extraction, aesthetic flattening, and environmental cost. He quotes Marx—capitalism as vampirism—to argue AI resembles a system that feeds on others’ labor. That’s sharp persuasion: use an authoritative frame to make the point. It confirms suspicions people already hold and spots where the technology functions as a value-extractor.
He also points to aesthetic problems. AI tries for photorealism but often produces uncanny errors. Jude didn’t remove those errors. He kept them and re-labeled some of that ugliness as “poetry.” That reframing is an aggressive rhetorical move: convert the problem into an aesthetic strategy. It will comfort some viewers and infuriate others.
Then there’s the environmental argument. Training large models consumes energy. Jude does not dismiss that; he notes it as real. That admission adds credibility. He’s not brushing off the critics. He’s weighing trade-offs and choosing a path anyway.
John Cage, “no bad sounds,” and the art of accepting the ugly
Jude invokes John Cage: nothing is a bad sound; everything can become art. Apply that logic to images and you open the door to everything: Renaissance painting and TikTok clips, cinema classics and run-of-the-mill AI output. Jude refuses to place everything on the same plane, but he argues for inspecting everything. That stance irritates purists who want rules about craft, provenance, and labor.
Here’s the mirror: Jude says accept the ugly. You might respond: accept the theft? He hears that objection. He also hears the fear about jobs and commercial collapse. That empathy matters. He’s not blind to the fear; he’s choosing to experiment anyway.
Industry reaction and the moral economy of culture
The film world’s reaction breaks down along predictable lines. Festival circuits and art-house audiences, used to provocation, will debate the film as aesthetic and political provocation. Writers, actors, and guilds in established industries see a threat. When creatives are paid by the hour, the idea of algorithms generating imagery from scraped human work looks like theft, plain and simple.
Social proof is clear: Jude is an award-winning director, and that lends authority to his experiment. But authority is not immunity. The more authority you claim, the more attention your contradictions get. Jude’s Golden Bear helps him get the screening; it doesn’t shield him from labor arguments.
Can satire bite the hand that feeds it?
Here’s the tough truth: satire that uses its object risks being read as complicity. That does not mean satire should never use the object. It means you must accept that your satire will be read multiple ways. If your goal is to make people uneasy, to reveal hypocrisy, then using the tool may be the fastest way to show what the tool does. If your goal is to prevent the tool’s spread or to defend jobs, using the tool complicates your position.
Jude chooses complication. He will provoke and accept the backlash. That choice forces us to ask tougher questions about intent, effect, and responsibility. Are you calling him hypocrite, provocateur, or honest experimenter? Which label fits your priorities?
Where this leaves filmmakers, artists, and audiences
For filmmakers: ask what you want from technology. Do you want cheaper visuals, a way to prototype, or a production method that cuts whole departments? Each choice has costs. Use the word “No” to set your boundaries: No, I won’t let my work be scraped without consent. No, I won’t replace labor that needs fair pay. No, I won’t pretend there are no environmental costs. Setting boundaries invites discussion — not stone-walling.
For artists: consider the John Cage argument and the Marx one together. You can accept ugly outputs artistically while resisting economic extraction politically. That is a credible, consistent position. Commitment and consistency matter: if you choose to work with AI, say how and why. If you refuse it, explain the standards you protect.
For audiences: ask whether critique by use convinces you. Do you trust the artist’s intention? Does the result illuminate the technology’s harms or soften them? These are open questions—what do you think?
Jude’s next steps and what they tell us
Jude says he is making a film with no AI and finishing a short that uses an AI voice reciting Dante. That mixed approach matters. It signals he is not dogmatic. He will use a tool when it serves his vision and skip it when it doesn’t. That stance respects nuance and avoids ideological purity tests. It also raises a predictable response: some will call that opportunism; others will call it practical honesty.
Final assessment: a provocation that forces choices
This film is an argument in image form. It argues that AI makes ugly, “gross and slimy” things and that those things can become material for art. It also admits the technology’s economic and environmental threats. Jude’s choice to use AI while criticizing it is not an evasion; it is part of his tactic. He wants the contradiction to be visible and noisy. That will annoy purists, please some critics, and force a real debate in public.
Where do you stand? Do you think exposing the machine’s errors is a valid aesthetic move or a dangerous normalization? How does your answer change if the people whose work trained those algorithms were paid and acknowledged? What trade-offs would you accept: lower budget and new imagery versus lost jobs and scraped labor? These are the questions Jude wants us to discuss.
#RaduJude #Dracula #AIpoetry #FilmDebate #GoldenBear #AIethics #CinemaVsAI
Featured Image courtesy of Unsplash and Max Nüstedt (1vc27E78mpc)