NextFin

DexRobot CEO Says Dexterous Hands Are Key to Giving Humanoid Robots Real Intelligence

Summarized by NextFin AI
  • Humanoid robots require dexterous hands for practical applications, as emphasized by Zhou Chen, CEO of DexRobot, at CES 2026.
  • Biological evolution shows that hands precede intelligence, indicating that without advanced manipulation capabilities, robots will remain limited in functionality.
  • DexRobot is focusing on the integration of AI and hardware, specifically the interface for physical interaction, which is crucial for future robotics.
  • The company is committed to long-term investment in solving the technical challenges of robotic manipulation, aiming to enable robots to work effectively alongside humans.

Humanoid robots will remain little more than “dancing toys” unless they are equipped with truly dexterous hands, underscoring the importance of fine motor control in the next phase of embodied artificial intelligence, said Zhou Chen, chief executive officer of Chinese robotics startup DexRobot.

Speaking at the Talk to the World forum hosted by TMTPost on the sidelines of CES 2026 in Las Vegas from Jan.6 to Jan.8, Zhou said that biological evolution offers a clear lesson for robotics developers.

“Sixty million years of evolution tell us that hands are the foundation of intelligence — hands came before the brain,” Zhou said. “Without dexterous hands, humanoid robots will forever be toys that can dance, but not machines with real utility or ‘soul’.”

Zhou said DexRobot is focusing on what he called the “last millimeter” of AI and hardware integration — the interface between intelligent software and physical interaction with the real world.

“The hand is both the technical bottleneck and the key to unlocking the future,” he said. “It is where perception, decision-making and action come together.”

DexRobot is developing robotic hands and manipulation systems designed to allow humanoid robots to perform complex, delicate and adaptive tasks, ranging from industrial assembly to service and healthcare applications.

Zhou said the company is prepared to invest over very long time horizons to solve the problem.

“We are willing to spend ten years, twenty years or even longer on this,” he said. “This is not a short-term race. It is a foundational challenge for the future of robotics.”

His comments reflect a broader shift in the robotics and AI industry toward embodied intelligence — systems that combine machine learning with physical interaction, enabling robots to operate safely and autonomously in human environments.

At CES 2026, major technology companies including Nvidia, AMD and several robotics startups showcased advances in physical AI, world models and agent-based systems, signaling growing interest in moving AI out of the digital realm and into the physical world.

Zhou said progress in large language models and perception has been rapid, but manipulation remains one of the hardest unsolved problems.

“Without hands that can feel, adapt and act precisely, robots cannot truly understand or change the world around them,” he said.

DexRobot believes breakthroughs in dexterous manipulation will be critical to enabling robots to work alongside humans in factories, homes and hospitals, and to unlocking commercial applications beyond demonstrations and entertainment.

“Only when robots can really use their hands,” Zhou said, “will they stop being performers — and start becoming partners.”

Explore more exclusive insights at nextfin.ai.

Insights

What are dexterous hands in the context of humanoid robots?

What role does fine motor control play in embodied artificial intelligence?

How has biological evolution influenced robotics development according to Zhou Chen?

What are the key challenges in integrating AI with physical interaction in robotics?

What advancements in AI and robotics were showcased at CES 2026?

What is the significance of the 'last millimeter' in AI and hardware integration?

How does DexRobot plan to address the manipulation challenges in robotics?

What are the potential applications of dexterous humanoid robots in various industries?

What are the current trends in the robotics and AI industry related to embodied intelligence?

What limitations do current humanoid robots face without dexterous hands?

How does Zhou Chen envision the future of humanoid robots in human environments?

What long-term investment strategies is DexRobot pursuing for robotic development?

In what ways can dexterous manipulation enhance human-robot collaboration?

What distinguishes embodied intelligence from traditional AI models?

What are the core difficulties faced by robotics developers in creating dexterous hands?

How might advances in large language models impact robotics development?

What is the connection between perception and dexterous manipulation in robotics?

How do major tech companies view the future of physical AI and robotics?

What factors could limit the commercialization of humanoid robots?

Search
NextFinNextFin
NextFin.Al
No Noise, only Signal.
Open App