Cursor for Filmmakers
I have been thinking a lot about relational databases and the future of filmmaking. Presumably, so has Larry Ellison.
AI deals. Media consolidation. All signs point toward an industry-wide scramble for a technological, data-driven solution to rein in ballooning spending. As studios search for that big technological blockbuster, pouring money and data into AI, they seem to have forgotten the most important part of their data pipeline: the people who actually make their movies.
In all the speculation about how AI will change the media landscape, from user-generated AI content to cheaper VFX pipelines, I seldom hear an articulated vision about how any of these companies plan to empower their creators through this revolutionary technology.
Hollywood was notoriously slow to adopt the kinds of data practices Silicon Valley has refined over the last few decades. But eventually, they got the message. Studios now have extraordinary visibility into performance, spend, audience behavior and risk.
Filmmakers rarely get access to that same level of tooling and visibility. Most work inside a constellation of disconnected tools that are excellent at producing artifacts, but poor at helping someone understand their own work as a living system. Context is scattered. Decisions are implicit. Creative authority is often procedural rather than expressive.
This isn’t just an artistic problem. It’s an economic one.
When creative intent isn’t legible, iteration slows down. Rework increases. The only lever left to pull is scale: more money, more data, more automation.
For filmmakers, the increasing scale and complexity of modern productions is a difficult challenge to overcome. Months and millions of dollars between intent and asset.
If you, dear reader, happen to be a venture capitalist, you’ve likely been hearing a lot of Cursor for X pitches. It seems like anyone who is anyone is getting one. Cursor for lawyers. Cursor for designers. Cursor for municipal bus mechanics.
For the uninitiated, Cursor is an AI-powered code editor. But unlike a simple chatbot that answers questions in a vacuum, this product class has context awareness.
For the user, this means that your editor can now interact with and understand your codebase. It can translate your intent into execution while considering wider implications across your codebase. For the company itself (and its VC backers), the real value is the data collected from being fully embedded into the lifecycle of an idea through its execution. The bet is that this data will be highly valuable in training new, efficient and specialized models, continually improving the product and empowering software engineers.
So, what does an application that empowers filmmakers actually look like?
I’m envisioning an application for writers and directors, where story exists in multiple modalities. A new paradigm for screenwriting and pre-production. Jump between writing in screenplay format to “writing” through generative storyboards or quickly captured reference video.
This isn’t an AI movie generator. It’s an experiment in what happens when you treat filmmaking like an intelligent living database rather than a collection of static versioned assets. A place where creative artifacts are relational, navigable, and meaningful across time. A creative decision in an edit will open a PR to update the script. A change in the script will auto-create and, if so desired, auto-complete a series of downstream tasks.
The next era of filmmaking, just as every era that’s preceded it, will be defined by the hardworking, creative and inspired people who make the movies we love. I want to build for them.
Noise Floor is a blog about creative technology and media production. This post is the beginning of a series where I will be exploring new filmmaking workflows through building in public. Subscribe to follow along!