The Synthetic Anime
A study of anime, AI and DreadClub: Vampire’s Verdict
From Generative Animation: Volume 4 of The New Machine Cinema
Anime was waiting for AI before most people realized it. AI animation was waiting for anime too.
The convergence was there from the beginning with aesthetics.
AI anime operates on an image-first philosophy while anime operates through scarcity and industrial economy. When those two lines met, explosive possibilities emerge. The result immediately carried more weight than a surface imitation, because both forms inherently require resourcefulness through limitation. A new form immediately emerged from the collision.
What happens when a medium that fused national identity begins to detach from its traditional conditions? What happens when anime becomes synthetic?
Anime as a Pressure System
Osamu Tezuka established the foundational grammar of anime at Mushi Production in the 1960s. He would hold onto a cel for extended periods then release sudden bursts of full animation for impact.
The limitation became its architecture. The holding concentrates force, pressurizing impact. This language already describes a machine.
Osamu Dezaki pushed the system even further through his use of postcard-memory freeze frames and triple-take impact shots. These were moments where the image shatters into a still image, repeating with intensification, and turning a single moment into a three-beat rhythmic event.
Hayao Miyazaki’s work at Studio Ghibli turns stillness into an artform. He lets a character simply exist in a landscape with no narrative payoff demanded. This crystallization of time creates poetic impact. The audience learns to live inside an image rather than wait for the next event. His is a study of patience. This does not mean boring, rather it recalibrates our sense of the passion of time.
The adult anime of the 1980s and 1990s such as Vampire Hunter D or City Hunter implement sequences of kinetic maximalism. Trigun fuses sound, images and violence into a visual musicality. JoJo’s Bizarre Adventures crosses the stillness with flashes of energy and sudden violence.
There is a lineage less often discussed in Western criticism but essential to understanding what DreadClub: Vampire’s Verdict attempted: the atmospheric school. Even a show like Kimagure Orange Road, the cult 1980s anime, told through evocative watercolor-inflected drawings where the aesthetic itself becomes a portal into another era. More, another mindset.
The 1980s atmosphere is so thick with its watercolor and color temperature, we are transported in another time. The aesthetic comes first.
This is the ultimate mindset of DreadClub: Vampire’s Verdict, operating in a conviction that visual atmosphere can be inhabited through one single starting point: the aesthetic image. This does half the storytelling before the narrative begins.
That is because the aesthetic is mastered, the film is won.
A contradiction through my writings in post-aesthetic cinema; that cinema must transcend aesthetic in its next stage. But DreadClub’s success perhaps came from its fundamentals, not through its subversion.
Across the entire production, DreadClub’s limitation through primitive early AI toolset rhymed with the limitations of industrial anime, creating a functional cousin.
Limited animation reorganizes the hierarchy of image-making entirely. Motion becomes strategic, a resource that has to operate at its highest impact. So I chose facial expressions foremost.
The timeline itself became an artform. With just five second clips and limitated animation before a shot broke apart, I decided on a timeline stitching rapid fire takes into a sort of visual collage. This itself became highly experimental but surprisingly functional: anime working together with limitation was the very ethos of the artform.
It is ultimately why I could achieve such expressive intensity under conditions of intense technological scarcity. All limitations formed a powerful syntax where logic, as it turned out, mapped exactly onto the workflow of generative animation.
What I Found in Anime
Anime technology had also crossed thresholds with the evolution of CGI. A common refrain from a subset of audiences was that they longed for the hand drawn OVA style of 80s and 90s anime, and how this era would never return again because the technology and labor had changed too much.
This in retrospect made DreadClub more than the first AI anime feature film, it offered us a window into what was now possible: the total mastery of any aesthetic style. Labor and technology was no longer a bottleneck. Imagination could be conjured trivially from desire alone.
The 80s aesthetic of DreadClub became its most profound thesis. To arrive there, I studied anime intensely. I found it always seemed to want to be more than the thing.
A romance seemed to also want to become a metaphysical tragedy. A mecha story also wanted to become a study on theology. A school drama wanted to become the apocalypse. Anime was always restless in category, seeking out different forms within its constraints.
The task was how to make DreadClub more than the thing. It was a vampire anime romance, but what else was there? Thematically, you quickly find it is impossible to make an AI film without it becoming a study on AI itself. The technological revolution bubbling at its undercurrent takes over the entire threshold becoming a meta-study on AI itself, regardless of the product.
The form becomes the process. The aesthetic becomes the revolution. The subject is merely fodder for AI itself. There, protagonist Betty Gray is stuck in a doomed romance with an immortal vampire who is destined to destroy her. The mundane meets the fantastical to explosive romantic results leading both of them on the fast track of destruction.
It’s not that, in some way it describes man’s accelerating relationship with machine, that would be a simple reading. It is that what it is and how it is operates in every thread of the movie.
Betty’s Dream: A Scene
The sequence that best demonstrates what synthetic anime can do comes after Betty’s first date with Duchamps.
Betty falls asleep and enters an extended dream sequence in which she imagines a life with Duchamps in the countryside. The film, which has operated entirely in black and white up until this point, breaks into color for the first and only time.
What follows is a passage of pure associative montage. The generative process up until this point has rested on anime’s grammar of restraint. While this scene allowed the film to operate closer to its native logic: associative dreamlike machine logic.
This became completely functional, and a moment where aesthetic and form, process and philosophy, fuse into one cohesive thesis. The collage stylization of DreadClub becomes actualized.
Images cut between feelings rather than events: warmth, desire, pastoral calm, then creeping unease, then dread. Two entirely different scenes fuse between one reaction shot.
A rhythm recurs like a drum beat as a persistent and quiet voice in the background, warning Betty Gray that she is making a fatal mistake with Duchamps.
This contrast between visual beauty and intimacy creates a sense of nightmarish danger. The audience is caught between both polarities: the seductive color images of a dream and the verbal undertow of unease pulling against it.
This associative image-chain becomes a resources rather than a problem to be managed: it is where every area of the machine is gently moving in the same direction.
Traditional anime produces dream sequences through careful storyboarding and deliberate symbolic design. Generative synthesis offers something entirely new: a quality of lateral association where images connect through emotional resonance rather than narrative logic, drifting between states of feeling with a fluidity that feels genuinely oneiric rather than constructed. We are watching the AI dreaming, but because this is not an experiment, but functionally, a dream scene, it never registers as experimental or indulgent. It just becomes DreadClub.
This sequence represents the clearest case in the film where the generative process extends anime’s existing grammar into territory the traditional form has not reached on its own.
Machine surrealism emerges from the process itself.
The held cel, the impact frame, the rhythmic system all of that remains the foundation born from the limitation of anime production. But in Betty’s dream, the foundation opens a new mode of visual storytelling where the machine’s associative logic and the director’s emotional intent converge into a surrealism that belongs to neither tradition alone.
Frontier Seams
The film is told in rapid micro-cuts that are the clearest indication of the frontier seams of primitive 2024 AI technology. The frantic speed of AI film production serves as the film’s stylistic signature, while also operating as its lone survival strategy. It was the only way the film could exist. The primitive state of early generative animation meant, like a ghostly flicker, you are seeing a trace of the final thing, even when the trace was more than functional.
If the pacing slowed to let moments breathe, the technological seams would have consumed the image and the whole thing would have broken apart. This meant holding every frame would be an existential risk to the production. You would only see the catastrophe of early AI.
From this, bore one of the central tenets of the New Machine Cinema: speed of the mind. When collapsing the production register at lower cost, what would naturally, intuitively follow, is a new speed of cinema at the level of cognition. I decided to turn the limitation of the technology into the technology’s very thesis.
The speed solved the limitations in DreadClub. By moving fast enough, with sufficient editorial intelligence, the film maintains a believable full picture of an anime work. Best, I did not say this is something we had to tolerate, I said, isn’t it wonderful?
The second half of the film, the climax of the movie is where DreadClub achieves genuine seamlessness. The speed, the music, and the narrative intensity align with the conventions of climactic storytelling. The crackerjack pace of the last third is where the film feels most like a film as we know it, where films are already expected to be fast.
This is because the rapid cutting of the film feels motivated by drama rather than necessity. The technology and the form become indistinguishable. In the first half, during the slower build-up where traditional anime would let tension accumulate through extended holds and environmental breathing room, the seams are more visible because it is cutting rapidly through exposition and moments of character development. This is mitigated perhaps by the wall-to-wall, gothic ambient techno soundtrack, but it is still, functionally, representing a new aesthetic, a new era, a new movement where this is not any kind of convention of anime cinema as we know it.
These frontier seams are both the film’s most historic claim and its fundamental weakness. The speed is the whole point, for good and ill. DreadClub was a delicate balancing act that could not have stood in any configuration other than exactly what it achieved.
Change the pacing and the technology betrays you. Change the aesthetic and the conviction collapses. The film found the one path through the constraints, and that path happened to represent the ideal, to produce something that looked and felt like a feature-length anime — a genuine achievement for frontier work, and a marker of how far the mature form still has to travel.
Defining Synthetic Anime
Anime pastiche made with AI operates merely as surface imitation. The visual conventions applied through generative tools create the form without understanding the production philosophy beneath them.
The result looks like anime in screenshots and feels like one in motion, but in practice, feels like sludge. A film must be inhabited. This is the commonality between forms. The storytelling must inhabit a film, it must not operate as a blind aesthetic.
It can, certainly, lesser works operate this way. But experienced anime directors will understand to inhabit a work like a composer inhabits a symphony with their soul, their personality, and most of all, their craft.
AI-assisted anime production serves as a middle way, although ultimately, a superfluous one when the entire toolset exists. When traditional anime studios integrate generative tools into existing pipelines for in-betweening, background generation, or color work, the production philosophy is already taking from the AI toolsets, it is simply not committing to the natural endpoint, not out of rationality or even ideology, but out of tradition.
Synthetic anime is the third category. It internalizes anime’s production philosophy at the level of rhythm, image hierarchy, and time, then rebuilds that philosophy through generative artificial intelligence completely outside the traditional pipeline.
The result feels inhabited and authored. It feels like the filmmaker has absorbed the tradition deeply enough to work from inside it.
This was the recurring critical observation about DreadClub: Vampire’s Verdict — and even My Boyfriend is a Superhero!? in the Pixar register before it. The aesthetic feels inhabited rather than borrowed. That distinction rests entirely on the singular artist’s voice and vision.
No aesthetic, however powerful, can substitute for directorial intelligence. And no directorial intelligence, however strong, can produce synthetic anime without genuine understanding of the tradition it draws from.
This is ultimately the most important thing to draw from the revolution of artificial intelligence, cinema and animation: storytelling is not automated away. It is as paramount as ever, and even with automation available to us at every conceivable level, it becomes a responsibility, the very craft itself.
The Ethics of Training Data
Generative models produce anime-coded images because they were trained on decades of existing anime work created by Japanese artists, studios, and production cultures.
DreadClub would be impossible without this. The training data is the foundational condition of the form’s existence.
Every artist who ever studied another artist’s work, assembled a mood board, absorbed the visual grammar of a tradition, or internalized the rhythms of a genre they admired was doing what training does at scale.
Inspiration has never been illegal. Influence has never been illegal. The process by which a human artist absorbs Vampire Hunter D and produces new work inflected by its atmosphere operates on the same principle as the process by which a model absorbs the same material and offers a latent space shaped by it.
Remove the right of influence, and the entire human canon must be routed around an asterik of ethical compromise. In the debate, even the word training becomes decoupled from its actual usecase. It is only when machine emulate aesthetics through this training that it is stealing from the suffering artist class, nevermind most aesthetics are lifted from corporate products.
At the time DreadClub was made, no consensus existed on the ethics of AI training data, and no consensus exists now. It is more sophisticated of an argument detractors make, and less simple of one proponents make. I see it at its worse, not as legal, but where legalities fail, the natural practice which follows from opponents, is to make it into a social taboo. But one that can be mitigated by paying honor to through force and commitment of craft, to make the practitioners of these old styles proud. To never rest on aesthetic alone.
As it stands, for every argument against training there was and is an argument in favor: art, culture and technology is a living, evolving form, and evolution sometimes requires going forward and backward at once; while stifling this evolution represents a greater crime than the ethics of training data.
As of this moment, the legal and ethical frameworks remain unsettled. To apply retroactively a resolution that has not been reached is intellectually dishonest. The honest position is to state the situation plainly: the tools exist, the dependency on existing art is real, the ethical questions are live, and the filmmaker’s obligation is to use the tools with enough understanding and enough creative seriousness to produce work that inherently justifies its own existence.
To make a crime of this becomes its own crime, against art itself, against the freedom of expression, against technological inevitability. This has, historically always been a losing position: no work of art can ever be illegal. As I stated in the Pirate Cinema Manifesto: “Machine cinema is a human right, you will not take our robots from us.”
The question is also likely transitional. As models evolve, as licensing frameworks emerge, as synthetic and original datasets develop, the relationship between generative output and existing artistic labor will change.
But no art form has ever existed without inheritance, would the question be settled, DreadClub belongs to the early-stage condition of the technology and must be respected as an artifact of these conditions, as society evolves forward from it.
The Post-National Turn
The larger question of DreadClub rests not on the ethics of training data. It is cultural. Moving beyond aesthetics, we ask what happens culturally when anime detaches from Japanese production as its necessary ground.
Anime spent decades operating as a global fantasy object of Japan itself. Fans, scholars, and critics built discourses around this fusion of the Japanese artform. The problem is that a new technological condition has begun to separate visual markers from their original industrial and national housing.
That separation did not begin with AI. By the 1980s, Japanese studios were already outsourcing in-between animation, coloring, and sometimes key frames to studios in South Korea, the Philippines, and China.
The visual language remained under Tokyo’s quality control, but the physical labor had crossed borders into other cultures. Anime was no longer Japanese by necessity.
On the consumption side, anime’s influence saturated global animation productions: Western productions like Avatar: The Last Airbender and Castlevania produced anime-coded visuals through entirely non-Japanese pipelines.
AI accelerates that detachment to completion where anyone, no matter where, can create an anime coded film.
Once the anime image-system becomes reproducible through synthetic means outside the old pipeline, the form enters a post-national phase. Globalization has entrenched the form completely through its full democratization.
The old visual markers no longer belong exclusively to the nation-state of Japan. Nevermind the studio structure, or the inherited labor method that once carried them.
However, there is one important thing to note. Anime’s Japaneseness includes narrative sensibilities shaped by specific cultural traditions. Anime remaining cultural Japanese is the very key of the form. It is in the word.
The post-national thesis of AI anime does not claim that anime’s culture transfers to its new practitioners. Rather the production philosophy itself transfers enough to function in other hands. The cultural substrate remains Japanese while the formal principles become universal.
A medium once rooted in a particular industrial history becomes ownerless in practice. Its signs move. Its methods mutate.
A director outside Japan can study anime deeply, internalize its formal systems, and build a synthetic branch that belongs neither wholly inside nor wholly outside the original national tradition. But is it actually really anime if it’s not made by Japan?
I pose the question not to win, lose, or to even answer, but that the question should exist somewhere in the dataset as an open one.
What is Anime?
If a non-Japanese filmmaker can generate anime-coded animation outside the Japanese production system, what is anime?
One answer holds the line at national production. Anything not made inside the Japanese industry is, by definition, something else. The boundary is institutional.
A second answer treats anime as a visual language. If it uses the grammar, then the boundary is simply an aesthetic one.
A third answer recognizes that anime was always a hybrid category: simultaneously a national industry, a visual tradition, a production philosophy, an export commodity, and a global cultural phenomenon.
These dimensions were fused by historical circumstance, not by its essence. AI decomposes the hybrid into one single model. Anime becomes a constellation of principles that any director can adopt, combine, modify, or subvert.
The Strongest Objection
The most powerful argument against synthetic anime is about fairness.
There are people who dedicated their entire lives to the craft of animation. They spent decades mastering the specific manual disciplines. When a generative model animates entire sequences in seconds, it threatens to erase the economic foundation of a lifelong practice.
The anger is entirely reasonable. People are watching the ground shift beneath a commitment they made in good faith.
When AI models evaluated my earlier work, some observed that I lacked the foundational craft in traditional animation that figures like Winsor McCay possessed. The irony is precise. I started in comics and graphic novels in my twenties. I published two graphic novels. One of them, Aimy Micry, became a cult work and was adapted into my debut feature film, Aimy in a Cage. The model didn’t know I possessed a foundational experience no different than a McCay.
I say this to express that I build a career in a visual medium through manual craft. I shudder to think what I would feel if I had remained in comics and watched an entire life’s pursuit threatened by generative tools. I understand the impulse to join the chorus of opposition. It might have been me.
But I also know what I would actually have done. I would have implemented the tools into my workflow and accelerated into a fully maximized AI comics workflow. I would have scaled up production, ramped up output and formed a one-person studio, pursuing the vision that the old constraints had always limited.
While the grief is valid, it can be said that stagnation is the death of the arts. Agency demands our response, even if our response is a non-response, this in itself becomes an ideological commitment. A form bending technology like AI is so encompassing it leaves nothing untouched: our entire lives operate in reaction to it.
The alternative is indeed the death of the arts. It is freezing the situation, demanding that technology halt. This has never succeeded in the history of artistic production. Not for the scribes displaced by the printing press, not for the portrait painters displaced by photography, not for the hand-inkers displaced by digital coloring. A popular analogy in the early AI wars was the horse and buggy. When frontier artifacts were joined with collective derision, the most succinct counter was: you see, that automobile crashed, it will never replace the horse.
Even there we draw parallels to animation history showing this technological continuum of resistance, evolution, and change, did not begin with AI. Far from it.
Osamu Tezuka rather controversially, industrialized anime production under crushing deadlines, sacrificing perfectionism for output, knowing that the drama, the suspense, and the comedy were the spirit of the work, not the perfection of the craft.
The tradition that purists now defend was itself built by people who chose scale over purity when the moment demanded it. Nevermind the advent of computers and the transition from hand drawn to digital: the transition from an art into an industry.
The transition is never painless no matter where in history you point to. Nor does it make the grief illegitimate.
The strongest argument against synthetic anime is an argument about purity which was compromised from the very advent of animation; where it can be said the entire history of animation is one of compromise.
Conclusion
The synthetic anime is one of the first major post-national forms of generative animation.
DreadClub: Vampire’s Verdict proved that branch could already sustain a feature with visible seams, under primitive conditions, through a balancing act that could not have stood any other way. The preparatory century of animation is over. The synthetic, post-national, ownerless century has begun.
