ByteDance is developing a new AI model capable of real-time spatial video generation, with founder Zhang Yiming personally overseeing the project. The company aims to compete with Meta and Alphabet in the robotics and autonomous systems sectors. The model could be released as early as next month, though the timeline remains subject to change.
Built upon the existing Seedance video generation technology, the model allows users to create interactive virtual worlds for gaming, short dramas, and live streaming. It is designed to respond to voice commands or physical movements from users of Pico headsets. The system generates video at 20 frames per second with a latency of approximately 0.05 seconds.
The technology shifts the computational load of spatial content generation to the cloud, potentially reducing hardware requirements and costs for VR devices. By focusing on AI-driven content rather than just hardware specifications, ByteDance seeks to challenge Meta and Apple in the mixed reality market. Both competitors have invested heavily in their respective platforms but have yet to achieve mass-market adoption.
