Stardust Era's Chen Jianyu: VLA Is Not the Endgame, Embodied Brains Have Entered the World Model Era | WRC 2026

Chen Jianyu, founder of Stardust Era, proposed at the WRC 2026 conference that Vision-Language-Action (VLA) models are not the endpoint of embodied AI. He believes the industry is shifting from simple action execution to a cognition stage based on world models, with large model technology being the key driver of this transformation.

Traditional robots rely on preset paths and struggle to cope with complex and changing environments. After introducing world models, robots can understand physical laws and predict environmental changes like humans, thereby autonomously completing tasks in unstructured scenarios such as logistics and services. This is an important step for artificial general intelligence moving into physical space.

This trend will drive the robot industry to upgrade from "automation" to "intelligence." Enterprises need to increase investment in large model training, multimodal perception, and physical interaction algorithms to build embodied brains with stronger generalization capabilities. This will also reshape the competitive landscape of future smart homes and industrial automation.

This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.