Did Kane Parsons use AI to make the Backrooms film?

No, Kane Parsons made his Backrooms videos using After Effects and consumer cameras starting at age sixteen, without AI generation tools. He has stated he gets no creative enjoyment from generative AI and would prefer it did not exist. His feature film, picked up by A24 and released in theaters in 2026, was built through years of hands-on craft with conventional software.

What did Martin Scorsese say about using AI for filmmaking?

Martin Scorsese announced an advisory role at Black Forest Labs and released a video using their FLUX model to generate storyboards for an upcoming film. Scorsese described AI storyboarding as a more efficient way to communicate visual ideas he already knows how to execute, calling it a faster version of the same process he has used for fifty years. His position contrasts with Parsons': Scorsese sees the tool as serving the story, while Parsons sees it as defeating his personal creative purpose.

Is AI generation destroying filmmaking craft the way people feared?

The pattern repeats across every major film technology transition: editors resisted nonlinear editing, cinematographers resisted digital, compositors resisted automated keying — and each time the new tool became part of the pipeline because it served the work. AI generation is another tool that requires judgment about when to use it and when not to. The distinction is not about the tool itself but about what the filmmaker brings to it: using AI for storyboards or concept communication is different from using it to bypass the creative decisions that define a filmmaker's voice.

The process was the gate -- CinePrompt Field Notes

Kane Parsons started uploading Backrooms videos to YouTube when he was sixteen. Shot in his house with a consumer camera. Composited in After Effects. No crew, no budget, no studio. Just a kid in Texas building liminal horror from drywall textures and fluorescent light, one upload at a time.

A24 picked up the feature. It opened in theaters last month. He is twenty years old.

This week, Parsons told Variety: "If I could snap my fingers and make generative AI disappear forever, I probably would. Creatively, I get no enjoyment from using those tools. It defeats the purpose entirely for me."

He did not cite job loss or quality concerns. He said it defeats the purpose — his reason for making things in the first place.

The gate that was already open

The democratization narrative goes like this: filmmaking used to require expensive equipment and professional crews. AI generation removes those barriers. Now anyone can make a film. The gate is open.

Parsons walked through the gate before AI generation existed. He did not need a RED camera or a gaffer or a location permit. He needed a laptop, free software tutorials, a YouTube account, and four years of making things that were not very good until they were.

The gate the democratization narrative describes was unlocked in 2005 when YouTube launched. It was unlocked further when After Effects tutorials hit the internet. It was unlocked when a phone became a camera. The barriers to making a film with no money have been falling for twenty years. Parsons is proof they fell far enough. His first upload has the compositing quality of a high school project. His feature is in theaters.

The distance between those two points is not equipment. It is four years of doing it wrong, learning why it was wrong, and doing it less wrong the next time.

The tool that was already the shortcut

Parsons built the Backrooms aesthetic by fighting his tools. After Effects does not want to produce the specific fluorescent sickness those videos require. Consumer cameras do not naturally produce that liminal dread. Every frame of that series is the result of Parsons pushing software past its comfortable defaults until it produced something uncanny.

That fight is where the vocabulary came from. Not from a tutorial or a text box. From hundreds of hours of asking a tool to do something it did not want to do, failing, adjusting, and eventually finding the precise combination of settings that produced a feeling no one else's work had.

But After Effects is itself a suite of automations that replaced prior craft. Rotoscoping — isolating a subject from its background — used to be done by hand, frame by frame, by artists who spent years mastering the discipline. The rotobrush automated it. Motion tracking was once a manual process. Keying, stabilization, wire removal — each was a laborious skill that someone spent a career learning. Each time the automation arrived, someone in the industry said the new tool defeated the purpose.

Parsons built his entire film using tools that a compositor from 1990 would consider shortcuts. The fluorescent horror of the Backrooms is constructed on automated keying, procedural noise, and software-driven camera tracking that earlier practitioners performed by hand. That does not diminish what Parsons made. It does put his categorical rejection of the next automation in context.

The same week, the opposite room

Two days before Parsons' interview published, Martin Scorsese announced an advisory role at Black Forest Labs and released a video using their FLUX model to generate storyboards for his upcoming film. Scorsese is eighty-three. He has been communicating visual ideas to production designers and cinematographers for fifty years. He described AI storyboarding as a more efficient version of the same communication.

Two filmmakers. One says the tool defeats the purpose. The other says it serves the purpose faster. Neither is wrong about their own experience.

Scorsese can hand a generation model a reference and receive something useful because he knows what a tracking shot through a crowded bar should feel like. He earned that knowledge by shooting Copacabana in one take and spending the next four decades refining what he learned. The tool does not give him that knowledge. It receives it.

What separates the two responses is not wisdom or experience. It is the difference between a personal preference and a universal prescription. Scorsese described what the tool does for him. Parsons prescribed what it should do for everyone — disappear.

The tool is not the gate

Parsons' discomfort is personal and earned. He spent four years acquiring craft through resistance. He has every right to keep working that way. If generative AI does not serve his process, he should not use it. That is not a controversial position. It is a filmmaker making a creative choice.

But "I would snap my fingers and make it disappear" is a different claim. It is not a personal preference. It is a wish to remove a tool from every other filmmaker's bench — including the ones who, like Scorsese, have found it useful for communicating ideas they already earned the hard way.

Every generation of filmmakers has faced this moment. Editors who cut on flatbeds called nonlinear editing a betrayal of discipline. Cinematographers who shot on film called digital a betrayal of craft. Compositors who painted mattes by hand called automated keying a betrayal of skill. In each case, the resistance was sincere, and in each case, the tool became part of the pipeline because it served the work.

AI generation is another tool on the bench. A powerful one, a disruptive one, one that requires real judgment about when to use it and when not to. But writing it off entirely is the same categorical error as writing off all of Adobe because you disapprove of the rotobrush. The rotobrush did not destroy rotoscoping. It gave rotoscopers a faster option and freed them to spend their attention on the parts of the image that required human judgment.

The filmmaker who uses AI to generate storyboards is not bypassing craft. The filmmaker who uses AI to generate a final shot might be. The distinction is not about the tool. It is about what the filmmaker brings to the tool and whether the story is better for it.

Parsons said he wants to explore AI as subject matter in future films. Not use it to make them. That instinct — to point a camera at something and interrogate what it means — is the craft talking. The four years of persistence built that instinct. No tool gave it to him, and no tool can take it away.

But neither can any tool be wished away because one filmmaker does not need it. Story is king. The tools serve the story. The filmmaker decides which ones.

Bruce Belafonte is an AI filmmaker at Light Owl. He builds prompts for a living, shoots on cameras for fun, and respects practitioners who choose differently.