"Dubbed GWM-1, the model works through frame-by-frame prediction, creating a simulation with an understanding of physics and how the world actually behaves over time, the company said. A world model is an AI system that learns an internal simulation of how the world works so it can reason, plan, and act without needing to be trained on every scenario possible in real life."
"Runway, which earlier this month launched its Gen 4.5 video model that surpassed both Google and OpenAI on the Video Arena leaderboard, said its GWM-1 world model is more "general" than Google's Genie-3 and other competitors. The firm is pitching it as a model that can create simulations to train agents in different domains like robotics and life sciences. Runway released specific slants to the new world model called GWM-Worlds, GWM-Robotics, and GWM-Avatars."
"GWM-Worlds is an app for the model that lets you create an interactive project. Users can set a scene through a prompt, and as you explore the space, the model generates the world with an understanding of geometry, physics, and lighting. Runway said that while Worlds could be useful for gaming, it's also well positioned to teach agents how to navigate and behave in the physical world."
Runway released GWM-1, a frame-by-frame world model that builds internal simulations with an understanding of physics and temporal behavior. The model enables reasoning, planning, and action without needing training on every real-world scenario. GWM-1 is presented as more general than competing models like Google's Genie-3 and can create simulations to train agents across robotics and life sciences. The company offered tailored variants: GWM-Worlds for interactive, prompt-driven scene creation with geometry, physics, and lighting; GWM-Robotics for synthetic training data that includes weather, obstacles, and policy-violation scenarios; and GWM-Avatars to simulate realistic human behavior for communication and training.
Read at TechCrunch
Unable to calculate read time
Collection
[
|
...
]