ByteDance Explores Autonomous Driving with Seed World Model Team
Decision Brief
According to 36Kr, ByteDance is preparing to enter autonomous driving, with the project led by the Seed world model team under Zhou Chang, focusing on unmanned logistics within Volcengine's auto business line. ByteDance has reached out to top AV talent and held discussions with some teams. However, ByteDance officially stated they are only conducting AI frontier research and have no plans for intelligent driving. Seed is ByteDance's Llarge model research team founded in 2023, with Zhou covering multimodal, world model, visual generation, and embodied intelligence. ByteDance has already entered the automotive cabin through Doubao large model, partnering with Seres for AIVA brand, but its ADAS partner is Yuanrong Qixing. Industry views suggest ByteDance's world model technology can integrate cabin and ADAS for a complete vehicle intelligence loop and accumulate data for embodied AI. ByteDance has ample compute and talent; although general world models differ from specialized AV models, they can quickly train specific models using traffic data and hire to fill engineering gaps. ByteDance's entry may disrupt the AV industry, especially teams using AI paradigms with converging product experiences.
Sources
- 36氪
Chinese tech/startup coverage; AI signal filtered from broader business news.
- 36氪
留言
登入后即可留言,和其他 builder 交换实测心得。
还没有留言,抢头香。