ACE Robotics: Embodied intelligence is likely to reach its 'ChatGPT moment' by the end of 2027, with data bottlenecks serving as a key variable.

Wang Xiaogang, chairman of Chinese embodied intelligence startup ACE Robotics (Daxiao Robotics), told Reuters in an interview: "We expect that by the end of next year, driven by world models and environmental data capture, embodied intelligence will usher in its ChatGPT moment." ACE Robotics, backed by Ant Group and SenseTime, raised over $100 million in the first half of 2026 and plans to launch an IPO "as soon as regulatory approval is obtained." Wang noted that current robots still struggle to reliably perform a wide range of tasks in unfamiliar environments, with the core constraint not being algorithms but a severe shortage of training data. "The entire industry has accumulated only around 100,000 hours of data over the past few years, which is far insufficient to train embodied foundation models." ACE Robotics plans to collect tens of millions of hours of real-world environmental data within two years and deploy its technology to 1,000 stores over the next year. At the industry level, Boston Dynamics released the mass-produced version of its Atlas robot in January, claiming AI has brought it closer to commercial deployment; Alibaba launched the Qwen-Robot Suite robot model kit in June, while researchers are also testing data collection schemes such as the HumanoidExo exoskeleton. The competition in embodied intelligence is shifting from hardware prototypes to a dual race in data scale and model capabilities.