arXiv 2606.19531 introduces ImageWAM, a "world action model" that uses image editing (rather than video generation) as the underlying primitive. The result: 1/6 the FLOPs of equivalent video-generation models, with comparable downstream task performance.
The "abandon video generation" angle: most world models for robotics and game AI are video-generation models — they predict the next frame. ImageWAM's insight: the "next frame" is mostly redundant with the "current frame" — only the changed regions matter. So instead of generating a full new frame, ImageWAM predicts the "edit map" (a sparse set of pixel changes) between the current and next frame.
The technical details: ImageWAM is trained on a large corpus of "frame pairs" (current frame + next frame), with the training objective being to predict the "edit map" — a low-resolution per-pixel change map. At inference, the model generates the edit map and applies it to the current frame to produce the next frame. The edit map is 1/16 the size of a full frame, so the FLOPs drop by 6×.
The benchmark: on a set of robotics and game-AI tasks, ImageWAM matches the performance of video-generation world models at 1/6 the compute. The "edit-based" representation also has a nice side effect: the model can be queried for "what would change if I do X" — i.e., counterfactual reasoning, which is hard for video-generation models.
The bigger takeaway: "world models don't have to be video models." The assumption that "world model = video generation" has been dominant, but ImageWAM shows that a sparse edit-based representation can be just as effective at 1/6 the cost. For the industry, this means world models for robotics and game AI can be deployed on much cheaper hardware, opening up new use cases.