TLDR: Wang Xingxing, CEO of Unitree Technology, asserts that current artificial intelligence capabilities are the primary bottleneck for mainstream humanoid robots. He anticipates a significant AI breakthrough within one to three years, which could enable these robots to perform complex tasks autonomously in dynamic, unfamiliar environments, akin to the transformative impact of ChatGPT on the AI landscape.
Hangzhou, China – Wang Xingxing, CEO of leading Chinese robotics firm Unitree Technology, has publicly stated that despite rapid advancements in hardware, the widespread adoption of humanoid robots remains significantly hampered by the limitations of current artificial intelligence technology. Speaking at the World Robots Conference in Beijing, Wang expressed optimism, predicting that the necessary AI breakthroughs could materialize within the next one to three years, enabling robots to operate autonomously in complex, unfamiliar settings.
Wang drew a parallel between the current state of humanoid robotics and the pre-ChatGPT era of AI, suggesting that the industry is on the cusp of a similar transformative moment. He emphasized that while robotic hardware is largely sufficient, the development of robust ’embodied intelligence models’ is the most critical task ahead for the sector. This sentiment was echoed across various discussions at the conference, which showcased rapidly advancing embodied intelligence technology, with humanoid robots taking center stage.
Chinese robotics companies, including Unitree, are actively trialing humanoid robots in industrial settings for tasks such as sorting and quality checks. Unitree recently unveiled its R1 humanoid robot, priced competitively at 39,900 yuan (approximately US$5,900), though it is not yet in mass production. Other firms, like Shenzhen-based Engine AI, are also planning to launch their own humanoid robots, with Engine AI’s model starting at around $5,400. This aggressive pricing strategy by Chinese manufacturers, significantly lower than Morgan Stanley’s projection of $200,000 in 2024 dropping to $50,000 by 2050, indicates a focus on market entry and volume, betting on future AI advancements to unlock broader utility.
Despite these efforts, real-world demonstrations in 2025 have highlighted the existing technological gap. Humanoid robots participating in events such as half-marathons and football matches often struggled, stumbled, or failed to complete their tasks, underscoring the current limitations in their intelligence layer. This gap between impressive mechanical capabilities and the required AI sophistication explains why companies can build advanced hardware but face challenges in achieving practical, widespread applications.
Also Read:
- NVIDIA Advances Robotics with New AI Models for Human-Like Reasoning
- AI’s Near-Term Workforce Impact and Its Effect on Tech Stock Valuations: An AInvest Analysis
China is intensifying its commitment to this cutting-edge industry, with substantial government support. The country’s robotics industry generated nearly 240 billion yuan (about 33.4 billion U.S. dollars) in revenue in 2024. Furthermore, China leads globally in humanoid robotics patent filings, with 5,688 patents compared to the U.S.’s 1,483. Established industrial robotics companies like Siasun and Dobot are also entering the humanoid space, injecting new momentum into the application of embodied intelligence technologies. The industry’s current strategy appears to be building a strong hardware foundation, anticipating that AI will eventually catch up to enable the full potential of humanoid robots.
