CEOInterviews.AI
Start App
Try: AI strategy tariffs earnings guidance layoffs interest rates growth

Wang Xingxing

Founder, Chairman, Chief Executive Officer and Chief Technology Officer, Unitree Robotics (Hangzhou Yushu Technology)

Search every verified Wang Xingxing interview, podcast appearance, and on-the-record quote — each transcript cross-checked by AI and human review to confirm speaker identity. Wang Xingxing, founder and CEO of Unitree Robotics, spoke at the 2026 World Robot Conference in Beijing on August 19, 2026. He discussed the current limitations of humanoid robots, stating that they lack sufficient efficiency and versatility for large-scale deployment in homes and factories. He noted that while robots can perform some basic assembly tasks, their efficiency remains lower than that of humans, and they require retraining for new tasks. Wang identified the generalization capability of embodied AI as a major global bottleneck, observing that success rates can drop dramatically when environments or objects change. He expressed hope that within two to three years, or possibly five to ten years, robots could complete roughly 80% of household tasks in unfamiliar environments, which he described as a potential "ChatGPT moment" for embodied intelligence. Wang also highlighted his company's work on enabling physical AI robot models to evolve themselves, using frontier AI foundation models to search for research papers and open-source solutions, write code, and generate robot control software. He mentioned a previewed robot developed in just over three months that reached a speed of 12.65 m/s and exceeded human jumping records. Additionally, he emphasized the importance of real-world data for AI training, suggesting that collaboration with local companies could help gather high-quality data. Wang reflected on Unitree's participation in the World Robot Conference since 2017 and 2018, describing the current period as a new starting point for accelerated robot evolution.

Recent appearances

  • FULL REMARKS: Unitree CEO Reveals Why Humanoid Robots Are Not Ready for Mass Deployment | AI1B

    Unitree CEO Wang Xingxing says humanoid robots remain too inefficient and inflexible for mass deployment in homes and factories. Speaking at Beijing’s World Robot Conference, he identified robots’ limited ability to adapt to unfamiliar tasks as a major global bottleneck. Wang predicted embodied AI could experience its “ChatGPT moment” once robots independently complete 80% of household tasks. For more details, watch our story and subscribe to our channel, DRM News. Unitree CEO Reveals Why Humanoid Robots Are Not Ready for Mass Deployment Unitree’s Wang Xingxing Says Robots Still Cannot Match…

    Watch on YouTube ↗

  • LIVE: Unitree CEO Wang Xingxing Speaks at Beijing World Robot Conference 2026 | APT

    Join this channel to get access to perks:    / @apt_news   LIVE FROM BEIJING — Unitree CEO Wang Xingxing speaks at the 2026 World Robot Conference in Beijing. Wang Xingxing, founder and CEO of Unitree Robotics, takes the stage as the global robotics industry gathers in China to showcase the latest advances in humanoid robots, AI-powered machines, intelligent mobility and next-generation automation. #Unitree #Robotics #HumanoidRobots #WorldRobotConference #AI

    Watch on YouTube ↗

  • Wang Xingxing of Unitree Robotics: Illuminating the Tree of Human Technology with AI + Robotics

    From the world’s first mass-produced quadruped robot to the humanoid robot that wowed the Spring Festival Gala, how does Unitree Robotics lead the era of mobile intelligent AI? Founder Wang Xingxing exclusively reveals the secrets of full industry chain independent R&D, the layout for a trillion-yuan market, and the path of technological disruption, taking you to witness the miracle of robots moving from laboratories to thousands of households. AI系列:MiniMax、智谱zhiupu.ai、宇树Unitree

    Watch on YouTube ↗

  • Wang Xingxing, Unitree CEO – 2025 World Robot Conference Keynote [English dub]

    Wang Xingxing, founder and CEO of Unitree, says the “ChatGPT moment” for robots is about 1–3 years away, though it could take 3–5 years if progress is slow. The biggest limitation is that large AI models for robotics are still not quite sufficient. He feels there is too much focus on data and not enough on model architecture, which remains not good enough and not unified. Current VLA models have poor real-world performance, and adding reinforcement learning is still not enough.

    Watch on YouTube ↗

  • Wang Xingxing: Unitree Robotics: let more people enjoy quadruped robots [ICRA 2020 WS Legged Robots]

    Speaker 7: Xingxing Wang - Founder & CEO, Unitree Robotics Title: Unitree Robotics: let more people enjoy quadruped robots Abstract: Xingxing will be talking about some new developments of Unitree Robotics. Bio: Wang Xingxing is the founder and CEO/CTO of Unitree Robotics. During his master's degree (2013~2016), he independently developed a low-cost high performance quadruped robot XDog (probably is the world's first 12 DOF low-cost high performance quadruped robot driven by out rotor BLDC motor). After graduated (May 2016), he joined DJI. In August 2016, he founded Unitree Robotics. Up to…

    Watch on YouTube ↗

The summary and quote tags on this profile are produced with AI assistance from verified, first-person interview transcripts, then checked by our team to confirm the speaker's identity and the accuracy of every quote. See how we verify →