27 October 2025
![]()
Unlock Infinite Advantages: Subscribe to Annual Membership
Image Credits: Visionaries Network
Taipei became the stage for two radically different yet strikingly complementary visions of the future of robotics at SEMICON Taiwan 2025. On one side stood Dr. Hiroshi Ishiguro, Japan’s most renowned humanoid robotics pioneer, who spoke less like an engineer and more like a philosopher. On the other hand, was Dr. Jwu-Sheng Hu, executive vice president of Taiwan’s Industrial Technology Research Institute, whose focus was firmly anchored in strategy, supply chains, and the economics of robotics. Together, their talks revealed both the dream and the discipline shaping tomorrow’s robotic society.
For Dr. Ishiguro, avatars are not simply machines; they are an extension of human identity. Defined as robotic or digital agents acting on human intent, avatars, he argued, represent the next wave of societal transformation, comparable in scale to semiconductors. The COVID-19 pandemic, by normalizing remote presence, provided the cultural breakthrough that avatars had long awaited.
His reasoning is simple yet profound: the most intuitive interface for humans is other humans. Humanoid robots, therefore, are not indulgent limitations but the natural medium for human-machine interaction. Ishiguro calls his work constructive science a way of building machines that teach us about ourselves, from cognition and emotion to the elusive phenomenon of consciousness.
Recent advances in large language models have solvd a problem that haunted robotics for decades: natural, humanlike conversation. He demonstrated this through his own robotic double an avatar that can conduct media interviews in multiple languages, including mandarin Chinese, which Ishiguro himself does not speak. At global expositions, he deploys dozens of robot guides, hinting at a future where physical presence is optional and agency is unbounded by biology.
Japan’s moonshot program embodies this ambition. Its goal for 2050 a society where humans are no longer constrained by their bodies, brains, or geography. The seeds are already sprouting.
Dr. Hu’s narrative, by contrast, was pragmatic. His focus was on the tectonic shioft from automation to autonomy. In the age of automation, humans are supplied both goals and instructions in autonomy. In the age of automation, humans supplied both goals and instructions; in autonomy, robots devise their own plans once given a target. AI has made this leap possible, enabling complex, AI has made this leap possible, enabling complex, multi-joint motions once thought infeasible.
But autonomy comes with hurdles: Overheating, heavy power demands, and above all, a scarcity of behavioral data to train what hu calls large behavior models. Two strategies are taking shape. Hu’s perspective also spotlighted Taiwan’s strategic position. strategic position. With its ecosystem of SMEs and manufacturing expertise, the island nation is emerging as a linchpin in the global robotics supply chain. The country’s roadmap emphasizes embedding AI into existing systems, advancing next-generation hardware, and nurturing startups. ITRI itself has spun off companies developing robotic exoskeletons and synthetic robotic skin. And under a new US$20 billion, four-year program, Taiwan will launch a National Robotics Centre to consolidate leadership.
For Hu, robotics is not about spectacle. It is about solving problems and taking on the “dirty and dangerous jobs that humans need not perform.
27 October 2025
27 October 2025
24 October 2025
23 October 2025
We offer automotive expertise for market projections and customizable research, adaptable to diverse strategic approaches.
Contact Us