ACE Robotics CEO: Robot Brains Will Have 'ChatGPT Moment' by End of 2027

Humanoid robot technology is on the cusp of a significant breakthrough, potentially mirroring the transformative impact of ChatGPT on artificial intelligence, according to Wang Xiaogang, CEO of Chinese embodied AI startup ACE Robotics.

Wang anticipates that robot brains will achieve this "ChatGPT moment" by late 2027. He explained that this advancement will be fueled by the development of world models and enhanced environmental data capture capabilities. However, he cautioned that widespread commercial adoption of these embodied world models across various sectors might take an additional four to five years after this critical inflection point.

While large language models like ChatGPT have become commonplace, AI systems enabling robots to navigate and perform tasks in unfamiliar physical environments remain a significant challenge. In China's burgeoning humanoid robot industry, investors are increasingly prioritizing real-world deployment over demonstrations like dancing or athletic feats, seeking tangible economic value.

“

We expect to reach the 'ChatGPT moment' for embodied intelligence by the end of next year, driven by world models and environmental data capture.

Embodied AI models are crucial for a robot's intelligence and autonomous operation. Unlike language models, these physical AI systems are designed to help robots perceive and interact with the real world in real time. This contrasts with predictions from other industry leaders, such as Wang Xingxing, founder of Unitree, who suggested a breakthrough could occur in two to three years.

ACE Robotics, founded in July 2025 and supported by Ant Group and SenseTime, has secured over $100 million in financing this year. The company is eyeing an initial public offering (IPO) as soon as regulations permit. Despite its relatively small 4 billion parameter count, ACE's open-source Kairos-4B model leads global public benchmarks, outperforming larger models from competitors like Nvidia and Ant Group. The company is actively addressing the data bottleneck, aiming to collect tens of millions of hours of data by using sensors on workers in production lines.

“

Even if we reach that inflection point by late 2027, it will likely take another four to five years to see broad commercial implementation of embodied world models across sectors.