When Everyone Can Make Films, What Makes Film Art?

The disappearance of constraint.

In 1976 making a feature film cost seven years and ten million dollars, minimum. You needed financing. You needed a studio. You needed a hundred people. The constraint sat in the economy. Who could pay? Who got a camera? Who got funding?

Now, 2024, you can generate a feature film in AI in two days. You type a prompt, you wait, you have video footage. Not perfect. But recognizable as film. Not to stream on Netflix, perhaps, but as document, as proof, as moment. It can happen.

This poses a question that film history has not had to ask. If everyone can make films, what is film? Not as medium (film is still film), but as practice, as stance. What distinguishes your film from my film?

Eighty years ago the answer would be: skill, taste, capital, connection. You had to be able to make films to be allowed to make films. Now you only need language. You must be able to write what you want to see, and the algorithm sees it.

This sounds democratizing. And to some extent it is. Many people who could never make film art can now make something that looks like film. But it also says something. If the constraint is gone, what remains? Only taste. Only choice. And that is much harder than technique.

In the past century, technical skill and artistic vision were inseparable. You could not make films without both. You had to be able to cut, to light, to concentrate in the moment because you only had one moment to take. That technical constraint forced refinement.

Now technique is no longer a constraint. Everyone can generate anything. So what remains is pure choice. What do you want to see? And that is much heavier. Because you cannot hide yourself behind difficulty. You cannot say: this is hard so it is good. Now everything is easy, so only taste counts.

This happened before in literature. With printing, writing became democratized. Everyone could publish. But almost none of what was published turned out to be art. Most was just spoken thoughts without form. The art lay not in publication. It lay in selection.

Films will probably take the same path. We will be flooded with AI-generated films. Much will be bad. Much will be contentless. And then we will have to discover: what makes a film art? Was it always technique that separated artwork from fake? Or was it something else?

My suspicion is that it was something else. That great films were never great because they were difficult. They were great because they carried choice. They said: I look at the world this way. And that stance, you cannot generate. You must take it yourself.

Sources: AI video generation studies 2024; film history (David Bordwell); production economics