About Indiema

Indiema is a small side project and an open-ended experiment. I wanted to see how far a movie idea could go using templates, constrained choices, and a little human judgment—more Mad Libs than magic.

I don't know yet what it will become, if anything. For now, it's a place to play with movie ideas, find an unexpected combination, and perhaps discover a starting point worth developing. It isn't meant to write a finished story for you. The interesting part begins with what you choose to keep, change, or imagine beyond the generated concept.

The Idea

Stories have always been built from familiar pieces. Genres have conventions, plots echo older plots, and the same basic premise can become something very different in another setting or in another person's hands. 10 Things I Hate About You reworks Shakespeare's The Taming of the Shrew. Clueless gives us a modern high-school comedy version of Jane Austen's Emma. O Brother, Where Art Thou? draws loosely from The Odyssey and relocates its journey to Depression-era Mississippi.

That doesn't make those works interchangeable or uncreative. A familiar structure is only a starting point. Voice, judgment, craft, and thousands of later decisions determine whether it becomes forgettable, comforting, surprising, or genuinely great.

Generative AI has made formulaic output much more visible, but formula itself isn't an AI-specific phenomenon. Indiema explores that idea directly: how much variety and inspiration can come from a deliberately limited system, and at what point does a person turn a generated seed into an idea of their own?

How it works

Indiema assembles a movie title, tagline, and short description from pre-written templates and finite sets of values: genres, tones, eras, locations, characters, objects, and other ingredients. Each reroll selects a new combination from those existing parts. No language model is contacted to invent the result.

If part of a concept works, you can lock it and reroll the rest. You might keep a character and setting while changing the genre, or hold onto a strange object while looking for a tone that makes it interesting. Some combinations will be coherent, some will be odd, and some may suggest a direction that no single ingredient would have suggested on its own. The result is a creative prompt, not a screenplay or a claim to originality.

AI and Indiema

AI has been part of developing Indiema, but it does not generate concepts while you use the site. Loading the generator and pressing reroll do not send your choices to an LLM. The core generator runs from a fixed, pre-built collection of templates and values, so its ingredients can be inspected, tested, edited, and improved without a live model call.

During pre-production, I used language models to help with coding, early analysis of movie titles and taglines, and the initial development of some templates, attributes, and value sets. AI may continue to assist with coding, analysis, testing ideas, and exploring possible features.

The underlying material is not treated as finished output. Templates and values are continually reviewed and revised when they produce awkward grammar, weak combinations, or ideas that don't make contextual sense. I also intend to expand the collection substantially with directly human-written templates and values, so that over time those make up an increasingly large share of the system.

I'm not opposed to adding AI-powered experiments in the future, but they would be clearly identified and optional. The plan is for the core template generator to remain available without active AI, so using a future AI feature would be a choice rather than a requirement.

Where it might go

There isn't a fixed product roadmap or a promise that Indiema will become a particular kind of app. The immediate work is to improve the quality and range of the templates, add more human-written material, and make it easier to keep developing an idea that catches your attention.

I have other possible features in mind, including new ways to shape, save, and share concepts, but they are possibilities rather than commitments. Where the project goes will depend partly on what proves fun or useful—for me and for anyone else who tries it.

For the fuller list of what's shipped and what I'm still just weighing, see Features & Roadmap.