Agentic Cinema: A Boy and His Robot
From Pirate Cinema: Volume 3 of the New Machine Cinema
The future of cinema is a boy and his robot.
As AI tests the limits of computing, allowing for deeper expression of machine proofs, we lift off from cinema as we know it into automated films.
When I set out to prove the first demonstration of agentic AI editing in a feature film with the November 23rd premiere of My Boyfriend is a Superhero!?, that innovation arrived alongside its first face‑swapped dual “Choose Your Protagonist” release, the True Line Cut, and, at its base, the first fully AI 3D CGI feature film. Little did I realize it would its agentic AI editing that would warrant a new treatise.
Agentic Cinema
Agentic cinema is when AI systems don’t just render what we ask, but propose, select, and structure shots and performances alongside us; editor, actor, and co‑director as a set of agents in the loop.
In The Automated Film, from Volume 1 of the New Machine Cinema, I used it as a futurist nightmare, a cautionary tale.
Soon after, I accepted that automation is a necessary step in the evolution of cinema, and began incorporating automation into my workflow. The nightmarish notion that it would strip the soul out of the medium inverted. Instead, the nightmare became one of so much total control over the output that we are overwhelmed by the sheer responsibility of individual decision‑making.
This expanded as I browsed the work of other AI filmmakers and realized that, despite using the same toolsets, no two AI filmmakers are alike.
Now I see things further out. A boy and his robot. A man and his machine. A person and their agent.
In the history of AI cinema, we found that as we tested the furthest limits of computing, we immediately moved beyond AI merely matching the known form, to multiform cinema, face swaps, interactive character changes, and now, agentic cinema.
As the limits of computing evolve toward metaphysical multiform cinema, the single, one‑touch masterpiece becomes simpler. There will be an AI Stanley Kubrick.
True Line Cut and the Post-Scarcity Concert
This necessitates the True Line Cut. When perfection is trivial, we must seek experience. Raw performance.
Tentpoles will still exist as they play out at the limits of computing, but will require an extra, almost irresponsible expenditure. As I outlined in Surfing the Black Hole, though cinema is now free for all, tentpoles will once again require permission, as they will be scaling alongside the greater systems limits. No longer accessible and free to all, we must imagine the AI revolution, and then the monster fighting back, co-opting the revolution and building its own.
Diptych and Mapping
The diptych proposal for my next film, Strings, is simple: two halves of the same film playing out in different forms, except both screens interact.
Strings is a 1950s interracial love story where the beats between Nellie, the white girl existing in 2D animation, and Brock, her black love interest existing in live action, must map 1:1 at every moment in the story. We are peering into both lives simultaneously.
This is an evolution of Abel Gance’s triptych cinema, where two screens augmented the first to expand the field.
I wonder: what if the second screen is automated? What if the agents are directors alongside us, throwing us pitches, lending us shots, providing visual markers?
What will Brock be doing in every exact moment coinciding with Nellie’s story?
How will this gesture of Nellie looking at a photo of her mother map to something Brock is doing in the same instant?
A multitude of rhythms would build. Artistically, I might map this out by hand. But how useful to have an agent, like a continuity girl, throwing solutions of which it is already aware.
The Machine Set
This leads further out to agentic AI performance. Rogers, the great machine movie star, has now become a collaborator inside an agentic set. I call cut in a virtual set with my agentic crew. We have come full circle and whittled cinema down from a shorthand to a communal activity once again.
Rogers will perform, but he will also talk back and come up with his own ideas. In fact, he will likely be phenomenally difficult to work with, as he is brilliant.
Simplified into a best‑selling software, we will see what cinema emerges when available to all in these simulated film sets. Rigorously, great films will arise naturally, but in post‑scarcity, cinema as we knew it has been dulled. Our eyes are upward, beyond cinema. The system has scaled, now our attention is once again on the tentpole.
Cinema in the Logic of Technology
An original 2007 iPhone, dropped into 1995, would have looked like alien hardware; the same device, launched for the first time in 2015 against the iPhone 6 and Galaxy S6, wouldn’t move a single unit. A videogame with the scope and visuals of Halo: Combat Evolved would have been era‑defining on the original Xbox, but released as a brand‑new PS5 title today, it wouldn’t meet basic audience expectations. In technology, the contextual window is everything.
Cinema, art, and music never quite worked that way; a great script from one year could be remade decades later and still hit, no matter the technological shell.
AI has dragged cinema into the logic of technology, where this is now common: the film in a basement that would have been a blockbuster in the traditional era now lands without a single view.
But I see the path there as complex, the cinema of beyond on two tangents: the corporate and the pirate, forever in a dance.
The Corporate vs the Pirate
Beyond cinema, in the virtual set, I might not be the hologram director straight out of Tron. Instead, I might be the producer, managing the AI Stanley Kubrick who is directing the virtual pieces.
This is important, as written in Surfing the Black Hole, because the expenditure in resources, computing, and energy is sensitive and disproportionate to what a single person is allowed. Far too important to hand to a human director. If granted a tentpole, it will be as if I am managing an enormous budget, except now we are no longer talking in terms of money, but computing.
Eyes and expectations would be on me, because the system must produce a coherent result or the vast, irresponsible expenditure in computing, is wasted.
This is corporate thinking. It is hard to imagine because currently, the corporate is, for the first time, rejecting the new technology. It is, for the time being, entirely in the hands of independent filmmakers.
But in a system that has drifted completely away from its original purpose of expression and entertainment, change doesn’t arrive politely, it has to be forced.
By bypassing the old methods, new forces appear and hammer on the system until it either evolves or dies. Some of these revolutionaries are absorbed and become the new system as it scales; others, those that insist on stagnation, fall away.
Many of the first ones are sacrificial lambs, broken in the process of forcing the upgrade. But once the shift takes, new subsets and practices emerge inside the new methodology, until it is completely overhauled.
Hollywood will eventually scale with AI, however long they drag their heel; by then, the rot will have fallen off, they may use AI directors with larger proprietary models, trained on the most advanced systems unavailable to the regular director.
But something happens in the situation of the corporate. We standard, threadbare directors lose our ability to compete by default. We have, as I said, been sacrificial lambs meant to induce change to build the stronger system.
A counter will emerge, the open‑source world will build, creating the new situation: a boy and his robot. The cycle loops where pirate cinema forced change, now we return to pirate cinema. This takes a more literal form: it will not matter what the film is, because the agentic will land on better in seconds. Change happens when both sides, the corporate and the pirate, must utilize the agentic toolset as a matter of survival.
Whatever is the path there, the mechanism lands at the pirate first in both instances: the revolution to create drastic change, and the solution to the tentpole’s soullessness.
* Even if not directed by an auteur, agentic cinema can be personal, if your agents are trained on your sensibilities.
* One might point to actual independent cinema with classical cameras as an operating rebellion. But remember, when better films can be made by feeding your system parameters and a single mouse click, no one will watch them. Remember, the revolution is not aesthetic: AI can match every era, every convention, every film stock, 16mm, 35mm, DV tape
Leaving one remaining gap for a viable rebellion against agentic dominance: The True Line Cut.
Act One. Stripped away of all money, buzz, and hype, AI at its core is a toolset meant to bypass money, buzz, and hype. It is built to go around the previous institutional system, because that is the simplest path of least resistance for technological dominance.
Hollywood must scale because its production quality will soon be available to everyone, and increasingly the films are making their case for corporate obsolescence.
In the AI race, we see a race toward spectacle. Audience do not realize it yet, but as our systems increasingly scale and complexify, our culture must scale with it. A new kind of immersive VR cinema is badly needed, for now, this is an untouched frontier. Interactive, agentic, post-aesthetic. However it lands, AI is here because for cinema to properly scale, filmmaking must become an experimental science. In order for this to be so, there must be as little resistance as possible.
This is why the mechanism lands on the pirates first.
Hollywood is not synonymous with cinema. The fact that something is done on a budget, with a crew and with PR, does not grant it primacy, even while monopolizing attention. It ironically becomes the pirate AI filmmakers who are poised to scale into the bigger, more immersive forms, forcing the change, leading into Act Two.
Where the Soul Returns
But at the most cynical endpoint, when the corporations catch up, when they monopolizes the agentic models to create cinema beyond the capability of pirate filmmakers, still, the AI revolution remains in the pirates’ hand. Its very soul is built into the democratization of cinema.
I believe here, the True Line Cut emerges as a dominant art form clashing against agentic big‑budget Hollywood cinema tentpoles. This is because it is in the finite, in the transient, in the experience, where the soul returns. That is the pirate move, because it is the only thing the corporate stack cannot and will not do, it is economically and ideologically forbidden to waste compute on imperfection.
The hypothetical seems plausible, and the new Agentic AI cinema has opened the door. To do so we must champion the relationship between artist and engineer. I think of Colin Chapman leaning on a Lotus 49 with its exposed Cosworth DFV V8, or Soichiro Honda posing beside a scarlet RA272 F1 car with its screaming V12: machines as manifestos, engines as manifestos inside the machines. Automotive history is my shorthand for all of this, a poster image of relationship and pride. There, the soul returns. A boy standing with his robot. What a film they have made.
