Cristóbal Valenzuela


The Cost of a Frame

April 2026

You can look at the history of media through the cost of producing a credible frame. I call this The Cost of a Frame.

Not a good story. Not a good movie. Not a meaningful work of art. Just the frame itself. The thing on screen. The difference between something that feels like a home video and something that feels like it belongs inside a film, a streaming show, a commercial, or a music video. That’s The Cost of a Frame.

Under that lens, most media falls into four buckets:

1. Low production value, high cost
2. Low production value, low cost
3. High production value, high cost
4. High production value, low cost

A lot of media history is the movement from the first bucket to the fourth.

For a long time, making even bad media was expensive. Cameras were expensive. Film was expensive. Editing was expensive. Distribution was expensive. Most people had no practical way to make anything, let alone anything that looked good. Home movies existed, but they looked like home movies. They were personal, not cinematic.

Then cameras got cheap. Editing got cheap. Distribution got close to free. Smartphones gave everyone a camera, and the internet gave everyone a channel. This moved us into the second bucket: low production value, low cost.

That is the world of most internet video. It is the world of YouTube, TikTok, livestreams, vlogs, reaction videos, tutorials, memes, screenshots, podcasts, and infinite casual media. It is also Sturgeon’s Law: ninety percent of everything is crud. The internet did not make everything good. It made everything publishable.

But there was still one protected category: high production value, high cost.

That has been the dominant model for most of modern media. Movies, television, prestige streaming, advertising, music videos, visual effects, animation, branded content. The expensive frame remained expensive. Even when cameras got cheaper, the frame itself did not. A good camera does not give you production design. It does not give you lighting, locations, wardrobe, extras, props, actors, blocking, set decoration, continuity, effects, or time. A Soderbergh or a Scorsese frame does not look expensive only because of the camera. It looks expensive because of everything organized around the camera.

This is the last expensive layer.

At this point is pretty obvious AI has begun to basically collapsed that layer.

The important thing about generative media is not that it lets people make videos faster. Speed matters, but it is not the main thing. The deeper change is that it lets people create frames with production value detached from the traditional cost structure of production. Every frame is a painting and everyone can paint. You can now generate images and videos that borrow the visual language of expensive media without needing the machinery that used to produce that language.

That changes what kinds of stories can be made.

Not just big stories. Small ones. Stupid ones. Personal ones. Stories too specific, too local, too strange, too private, or too unserious to ever justify a crew, a budget, or a production process.

That is what makes the fourth bucket interesting: high production value, low cost.

The is YouTube-scale creation with cinematic-grade frames. It is the production value of expensive media applied to the subject matter of everyday life.

A blockbuster frame used to require a blockbuster premise, because the cost had to be justified. You did not spend hundreds of thousands of dollars making a cinematic sequence about two dogs in an apartment building that hate each other. You did not build a world, light a scene, design costumes, create tension, and render atmosphere for a joke between neighbors.

Now you can. And that is the point.

To test this, I made a video in around twenty minutes. My dog, Henry, an Irish Terrier, and my neighbor’s dog, Milhouse, a Corgi, cannot stand each other. They cannot see or smell each other without reacting. Whenever one notices the other, they bark like they are about to destroy each other.

So I made a dramatization of what I imagine goes through their minds when they sense each other.

That video does not matter because it is going to become a movie. It matters because it would never have been worth producing before. The subject is too small. The audience is too specific. The premise is too silly. But AI makes it possible to give that tiny personal moment the visual treatment of something much larger.

That is new.

Photography went through a version of this. For a long time, photography had friction. Film had limits. Cameras were separate objects. Developing photos took time and money. Then smartphones made photography ambient. Once the cost of taking a picture fell close to zero, the meaning of photography changed. People did not simply take cheaper versions of the same photos. They started photographing receipts, meals, outfits, parking spots, whiteboards, pets, faces, jokes, injuries, sunsets, strangers, and themselves a thousand times over.

The medium expanded because the cost collapsed. And now the same thing is starting to happen to moving images, but at the level of constructed reality rather than captured reality.

Smartphones made it cheap to capture what was in front of you. AI makes it cheap to construct what is not.

The internet gave everyone distribution. The smartphone gave everyone capture. AI gives everyone production value.

This does not mean everything becomes good. Most of it will be bad. Maybe even worse than before, because now the bad work will look expensive. Sturgeon’s Law does not go away. It scales.

When production value becomes abundant, the scarce thing moves somewhere else. Direction becomes scarce. Point of view becomes scarce. Story becomes scarce. The ability to decide what is worth making becomes more important when making is easy.

This is why low-budget films like El Mariachi from Robert Rodriguez with a budget of $7K or Darren Aronofsky’s Pi with a budget of $134K are still useful examples, but not because they prove that cheap work can look expensive. They prove something more important: production value was never the same thing as artistic value. A film can have limited resources and still matter if the idea, structure, energy, and execution are strong. Story remains the core. AI does not change that. It just changes who gets to attempt the frame.

In the old model, production value acted like a gate. It determined who could participate in certain kinds of visual storytelling. You could have a great idea, but if the idea required spectacle, scale, atmosphere, period detail, fantasy, visual effects, or cinematic polish, you needed money. A lot of money.

This is why the fourth bucket matters the most. High production value, low cost is a cultural change. It means the expensive frame is no longer reserved for institutions, studios, agencies, or large teams. It can be used for a personal memory, a joke, a dream, a fake commercial, a family story, a neighborhood myth, a political argument, a children’s bedtime story, or two dogs that hate each other. The cinematic frame is becoming casual. Like photography.

That’s how mediums change. They do not only change at the top, where the best artists make masterpieces. They change when ordinary people start using them for ordinary things.

The camera became world-changing not only because it made great photography possible, but because it made photography normal. The internet became world-changing not only because it created new media companies, but because publishing became normal. AI media becomes world-changing when production value becomes normal.

We are entering the fourth quadrant. The Cost of a Frame is collapsing.