Agibot Chief Scientist Rejects LLM Path for Robotics

Agibot chief scientist Luo Jianlan rejected copying the large language model development path for robotics, arguing that embodied intelligence requires high-quality, multi-scenario interaction data that remains scarce. Luo warned that many so-called embodied foundation models are merely mid-training or fine-tuning, and called for a shift toward system-level concerns like data, models, and infrastructure in China's robotics scene.

Industry context: For AI and robotics practitioners, progress in embodied systems depends less on simply scaling model size and more on building integrated, deployable data loops and shared standards. Reporting by Kr-Asia describes Luo Jianlan , chief scientist at Agibot and an associate professor at the Shanghai Innovation Institute, arguing that embodied intelligence cannot simply copy the development path of large language models LLMs . Kr-Asia reports Luo warned that many so-called embodied foundation models are closer to mid-training or fine-tuning because high-quality, multi-scenario robot interaction data, including failures and long-tail events, remains scarce. The article traces a recent shift in China's robotics scene away from body-design fixation toward system-level concerns: data, models, infrastructure, and the ability for those elements to reinforce one another in real-world deployment, per Kr-Asia.