China’s Famous Investor Speaks Out on Mass Exit: Is Embodied Intelligence an AI “Bubble”?

Recently, Zhu Xiaohu, the managing partner at GSR Ventures, caused widespread debate in
China’s investment circles due to his controversial decision to massively withdraw from investments
in humanoid robot startups. Zhu cited unclear commercialization pathways and a growing industry
consensus as his reasons for the mass exit, sparking intense discussions about whether the current
hype around embodied intelligence represents a financial bubble.

The Debate Sparked by Zhu Xiaohu

Zhu’s remarks quickly became the center of attention, described by many as a “withdrawal
declaration.” His move was met with mixed reactions-supporters acknowledged the rationality
behind avoiding overly speculative investments, while opponents accused him of short-sightedness.
Zhao Tongyang, CEO of Zhongqing Robotics, strongly criticized Zhu for dismissing long-term
potential due to current commercialization challenges, highlighting a fundamental divide between
cautious investors and ambitious entrepreneurs.

Zhang Ying from Matrix Partners China also criticized Zhu’s remarks, cautioning that such public
skepticism could damage industry confidence and hinder future investment opportunities. Zhu’s
action, specifically targeting early-stage investments like those in startups StarSea and Songyan
Power, was perceived by many as opportunistic and short-sighted.

A Market Gone Wild: Sky-High Valuations and Rapid Financing

The embodied intelligence and humanoid robotics sectors have recently witnessed unprecedented
investment activities. Startups, some just weeks old, have secured hundreds of millions in funding,
often without concrete products or commercialized technologies. Data from the first two months of
2025 revealed China’s embodied intelligence startups raised 4.45 billion yuan across 27
deals-nearly equaling the entire previous year.

This trend has obliterated the traditional concept of “angel rounds,” with startups routinely receiving
sums previously reserved for later-stage investments. Companies like Tashizhihang and Yuanli
Lingji raised over $100 million within days of inception, indicating investor FOMO (fear of missing
out) is rampant, fueled by a belief in the inevitability of embodied intelligence.

Investors Chasing a New Technological Revolution

Investors currently fall into three main categories: those scouting underground academic startups,
those consistently betting on hard-tech regardless of market fluctuations, and a growing group who
only recently noticed the rising momentum in humanoid robotics. Additionally, industry giants and
large venture capital funds are increasingly dominant, with top firms such as Sequoia, Hillhouse
Capital, Tencent, and industrial giants like Lenovo now frequent participants in early-stage deals.
The frantic pace of investment raises significant concerns about valuation sustainability and capital
efficiency. Zhu Xiaohu highlighted these concerns clearly, pointing out startups with massive
valuations but minimal revenues and significant cash burn rates. Specifically, Zhu revealed that one
humanoid robot startup, valued at $300 million during its Series B round, was generating only $8
million annually with a $27 million cash flow deficit. According to internal GSR Ventures research
cited by Zhu, merely 14% of their analyzed projects demonstrated quantifiable cost savings or
revenue growth.

Future Prospects: Real Revolution or Risky Bet?

Despite justified skepticism, many investors remain convinced of the long-term potential for
significant companies to emerge in embodied intelligence. Industry veterans like Li Liang of
Hillhouse Capital argue that China’s comprehensive manufacturing base and recent AI
advancements position it ideally for breakthroughs in humanoid robotics. They believe success
hinges on effectively merging software advancements with robust hardware, providing efficient and
practical solutions rather than mere technological showcases.

However, as valuations soar and market excitement continues to climb, the real challenge now shifts
to how these startups manage abundant capital. Investors must carefully differentiate between
startups genuinely capable of long-term innovation and those merely riding the wave of speculative
hype.

Now we can see, Zhu Xiaohu’s controversial stance has not only ignited crucial discussions on the
future and value of embodied intelligence but also revealed underlying tensions between immediate
financial rationality and long-term visionary investing. Whether embodied intelligence proves a
transformative sector or becomes another burst bubble will ultimately depend on the industry’s
ability to convert technological promise into sustainable, marketable solutions.