Inside NinjaPunk: A New Kind of Filmmaking

There’s a common misconception that generative AI filmmaking is as simple as typing a few text prompts and waiting for a movie to pop out. The reality? That couldn’t be further from the truth.

Creating cinematic stories requires a team of artists—people with bold imagination, a clear creative vision, and a deep love for the craft of filmmaking. With generative AI, it also requires a hard-earned technical skillset: merging dozens of tools, stitching together custom pipelines, and having the same patience traditional filmmaking requires—to iterate, refine, and reimagine. At the leading edge, you’re also kitbashing open-source models, writing custom code, and constantly pushing the technology to behave the way you need. The tools are evolving fast—but they’re not plug-and-play. Not if you want to captivate audiences.

At Promise, we’re developing a film called NinjaPunk—an original IP from our Chief Creative Officer and co-founder, Dave Clark. Set in the neon-lit streets of Los Angeles in 2065, the story follows a cybernetic ninja on a path of vengeance after his wife’s murder—only to uncover a hidden underworld where ancient powers collide with futuristic warfare. We’re starting with a short film to bring this world to life—and from there, we’ll expand into a full-length feature.

Watch the Sneak Peek

NinjaPunk is a hybrid production, powered by a lean team with an independent spirit: fast-moving, hands-on, and deeply collaborative. The short film blends GenAI and 3D-created cityscapes and sets, AI-generated characters, and performances from live actors and stunt performers layered into advanced AI/machine learning workflows.  Why bring in real people? Because we believe human expression is essential to storytelling. We're experimenting with a variety of techniques to bring human performance to life, including blending live action with generative methods to create the most compelling and engaging stories possible.

Watch the BTS

To create intricately choreographed group fight sequences, we partnered with stunt team veterans from some of Hollywood’s biggest action films. We captured all of the live performance sequences over three intense days on soundstages in Los Angeles.

Next, Rob Nederhorst—Promise’s Head of Creative Technology and an acclaimed VFX artist with credits including John Wick 3, The Last of Us, and Tomb Raider—led the blending of 3D VFX with motion capture from our stunt performers and live performances from our lead actors. Many of these techniques are well-established, but when supercharged by GenAI, allow us to move faster and more efficiently, manipulate camera angles however we want, and iterate in real time—bringing ambitious ideas to life in entirely new ways.

Today, we’re excited to give you an early glimpse of NinjaPunk—and show you the progress we’ve made as we invent, experiment, and redefine what’s possible.

We recently sat down with an award-winning director—someone new to GenAI—to show him how we’re making NinjaPunk and get his feedback. His reaction? “This looks like... filmmaking!”

Ultimately, his response gets to the heart of it: Generative AI is creating new opportunities for storytelling—another chapter in the ongoing creative evolution that’s been unfolding for decades. Just as visual effects reshaped how stories could be told, AI is opening new ways to dream, build, and connect.

It doesn’t replace the craft—it deepens it. And we’re just getting started.

-George, Jamie & Dave

Co-Founders of Promise