This Animation Startup Wants to Make It Easier to Tell Open-Ended Stories

Why It Matters

Cartwheel’s controllable AI pipeline could lower production costs while preserving artistic intent, reshaping how studios and indie creators produce animation at scale. It also positions the company to lead the shift toward interactive, real‑time storytelling in games and media.

Key Takeaways

• Cartwheel converts 2D video into editable 3D skeletons
• Founders bring OpenAI and Google expertise to AI animation
• Control layer lets artists edit lighting, camera, and poses post‑AI
• Aims for open‑ended storytelling where characters react in real time
• Targets 3D workflow adoption even for 2D output

Pulse Analysis

The animation industry has long wrestled with a data bottleneck: while text and image datasets are abundant, high‑quality 3D motion capture remains scarce. This scarcity forces most generative tools to produce flat video outputs that are difficult to modify, leading creators to repeatedly re‑prompt AI in hopes of a usable result. Cartwheel tackles the root problem by training models on curated motion data, enabling the system to infer a detailed 3D skeleton from a simple 2D clip. By doing so, the startup bridges the gap between raw visual input and the underlying biomechanics that drive realistic movement.

Cartwheel’s differentiator is its "control layer," a post‑generation interface that hands the creative reins back to the artist. Instead of accepting a finished animation as final, users can manipulate lighting, camera perspective, and individual joint positions, effectively turning the AI output into a malleable power tool rather than a black‑box replacement. This flexibility directly addresses the industry’s fear of homogenized content; when creators can easily customize each element, the risk of a uniform visual style diminishes. The platform also streamlines the workflow for studios, cutting down on manual rigging and key‑framing while preserving the nuanced artistic decisions that define high‑end productions.

Looking ahead, Cartwheel envisions a world of open‑ended storytelling where characters possess dynamic motion models that respond to narrative cues in real time. Such capability could revolutionize game development, virtual production, and social media content, allowing endless variations without the need for bespoke animation for every scenario. If the company’s three‑year hypothesis holds—ubiquitous 3D authoring even for 2D outputs—it could accelerate the industry’s migration toward immersive, interactive experiences, setting a new standard for how AI augments, rather than replaces, human creativity.