NEWS | ByteDance is building a new AI model that generates spatial video in real time, and company founder Zhang Yiming is personally steering the project. The system is reportedly built on Seedance, ByteDance's existing video generation technology, and could launch as soon as next month, though the timeline may still shift.
Zhang has reportedly pulled together teams from across ByteDance's business units and directed heavy AI computing resources toward the effort. The move fits a wider shift at the company toward generative AI, as it works to keep pace with domestic rivals like Alibaba and Moonshot AI while also competing with Tencent in enterprise AI software.
The new model is designed to let people build interactive three dimensional worlds for uses like livestreams, short form dramas and games. Much of the heavy computing would run in the cloud rather than on local hardware, which could lower the cost and power needed for devices like ByteDance's Pico headsets to run immersive content.
BMN's view is that this positions ByteDance to compete directly with Meta and Alphabet in the fast growing race to build so called world models, systems that simulate environments people can interact with. If ByteDance pulls this off, it could pressure other video and AI companies to speed up their own real time projects or risk falling behind on user engagement.
Could this be the moment spatial video moves from a lab experiment into something people use every day?
Your picture beside a comment comes from Gravatar, matched to the email you signed in with. Set one there and it appears here; without one you get your initials. How we handle this