AI should make stories cheaper to produce.

Bob Bering runs a studio and has an opinion about everything. He walks into the room certain that storytelling is a soft skill — the part of the business you tolerate on the way to the numbers.

Hana tells him about a woman running from a dinosaur. Through the forest, toward a shelter, hammering on a door to be let in. A few seconds later somebody knocks on the office door. The woman walks in, mud to the elbow. Then the walls come apart, and Bob is standing in the jungle of the story he thought he was only being told.

It is a fiction. It is also a fairly accurate picture of what is happening to this industry.

A storyteller imagines a world. A studio brings everything that turns imagination into something millions of people can watch: the capital, the crews, the cameras, the post, the distribution.

That second half cost a fortune, and nobody knew whether anyone would show up. So the studio carried the risk, and the studio kept the IP. The writer was paid — sometimes very well — but if the world became enormous, the upside belonged to whoever had financed its conversion into images.

You could tell that as a story about studios taking advantage of creators. It would be too easy. For most of a century the ability to produce was genuinely scarce, and whoever could put that much money on the table to find out whether a story mattered held most of the leverage, fairly.

Everyone talks about the cost of production falling. The more interesting consequence sits just behind it: AI mostly lowers the price of being wrong.

When finding out whether a story works takes three years and eight figures, you have to be right before you begin. So you commission the research, build the projections, convene the committee, and look for the thing that resembles the thing that already worked. That is not a failure of creative nerve. It is what happens when every bad decision costs a fortune.

A series that used to cost around two hundred thousand dollars and take months can now be made for a few thousand, in about a week. Once the first version is that cheap, you can stop arguing and start showing.

This is why short form interests us more than it probably should. Episodes of forty-five seconds to two minutes put a world in front of real viewers quickly, and then you watch what they do. Do they stay? Do they come back? Which character do they argue about? Do they want the next one?

A story can meet its audience long before anyone has committed the money to build its entire universe.

This is not a hypothesis. In 2024, China’s one-minute vertical dramas out-earned its entire theatrical box office — roughly seven billion dollars, at something like four hundred and seventy new dramas a day.

And the last gate is already open. The audience is holding the screen. You do not need a slot, a network, or anyone’s permission to find out whether your world has a public.

The analogy comes easily to us; part of this company comes out of venture.

A storyteller starts with a hypothesis: this world in my head can matter to other people. The first episodes exist to put that hypothesis in front of reality. An audience appears or it does not. If it appears, you invest more. If people come back, argue about the characters and wait for the next one, you start to discover what the IP is actually worth.

That makes the first season an MVP, and it makes the storyteller far closer to the founder of the IP than to a supplier delivering a commissioned script.

Our job is the rest of it: the capital, the production, the tooling and the distribution needed to get a world out of one person’s head and into a place an audience can walk into.

If the storyteller is genuinely the founder, the economics have to follow. That is not a new idea. Image Comics built a publishing house on it, handing creators infrastructure while they kept their characters. Different medium, same question: can you build powerful infrastructure around creators without needing to own what they make?

There is something strange about launching a studio with a film like this one. The first thing most people will look at is how good the AI images are. In a few years that will be the least interesting thing about it.

The models will improve. The images will improve. The edit will get faster. What took a full crew will take a few people, then almost anyone. What impresses you today becomes normal.

But somebody still has to decide that a woman is being chased by a dinosaur, that she knocks on an office door, and that the walls come down.

When everyone has extraordinary tools, having the tool stops being extraordinary. What stays rare is a story nobody else would have told this way: a character you are still thinking about the next morning, a culture, a joke, a strange taste, an obsession, a way of looking at the world too specific to dissolve into the average.