Galaxy General (银河通用) Founder Wang He: Robot Capabilities Will Bloom Over the Next Five Years

Forecast and why it matters

Wang He, founder of Chinese robotics firm Galaxy General (银河通用), told reporters that robot capabilities will “bloom” over the next five years as agent architectures and tool integration mature. The claim places robotics at the centre of the next AI inflection point: not just smarter chatbots, but systems that can plan, reach out to external services and complete real-world tasks. For Western readers: China’s tech ecosystem is already crowded with AI chip makers, cloud providers and robotics startups, and firms such as Galaxy General are positioning to commercialize integrated robot systems rather than standalone models.

A brutal market test: rapid AI shifts and failed bets

Why the urgency? Recent collapses in the AI startup ranks show how quickly competitive dynamics can flip. It has been reported that U.S. startup Yupp — which aimed to monetize human evaluations of large models and reportedly raised about $33 million in seed funding — announced it would cease operations less than a year after product launch. Analysts cite two problems: the industry’s fast pivot from single-model chatbots to multi-tool agents, and the higher-quality human feedback models now buy from specialist vendors. The Yupp case is not unique; high-profile projects such as OpenAI’s Sora and several boutique AI plays have been scaled back or shuttered amid soaring costs and shifting product requirements.

Geopolitics, chips and the path ahead

Geopolitical constraints also shape the race. U.S. export controls on advanced semiconductors have reportedly tightened access for some Chinese firms, pushing local companies to optimize software stacks and agent designs around available hardware. That dynamic amplifies Wang He’s point: even without parity in premium chips, rapid progress in algorithms, toolchains and software-defined robotics could deliver leapfrog gains in capability. Still, the sector is volatile — CAICT data show the global AI company count has exploded, but venture cycles and technical pivots mean many early entrants will not survive the next paradigm shift. The real question now is not whether robots will improve, but who will successfully translate agent-level advances into robust, affordable systems at scale.