Latest / Minds, Bodies, and Terawatts / Chinese Robotics Startups Dominate VC Funding

Transcript

Chinese Robotics Startups Dominate VC Funding · Minds, Bodies, and Terawatts

• 0:00
  [Host] This is Minds, Bodies, and Terawatts, where the machines get funded and the rest of us get nervous. Forbes reported on April 1st that Chinese robotics startups are dominating venture capital funding, with Fourier Intelligence raising $246 million to build general-purpose humanoid robots out of Shanghai.
• 0:16
  [Guest] Fourier's story is worth zooming in on. Founded by Gu Jie in 2015, this company started building exoskeletons for rehabilitation patients. Eleven years later, they've pivoted to full humanoid robots with a healthcare focus, backed by SoftBank, CICC, and Prosperity7.
• 0:30
  [Host] A medical device company building humanoid robots. Before we go further, there's a piece on unscarcity dot ai called "The Humanoid Robot Revolution: 50,000 Units Shipping in 2026" that puts this in context. Morgan Stanley just doubled its China humanoid forecast for this year to 28,000 units. That means more than half of all humanoid robots shipped globally in 2026 could come from Chinese manufacturers.
• 0:54
  [Guest] And Fourier is positioned to capture a chunk of that. Their pivot from exoskeletons isn't random. Exoskeletons teach you how joints move, how balance works, how to keep a human safe while a machine moves around them.
• 1:07
  [Host] So the rehabilitation work was basically a training ground for humanoid design?
• 1:11
  [Guest] Exactly. They spent nearly a decade learning biomechanics, then applied it to a full body. That's a pipeline most competitors don't have.
• 1:19
  [Host] And $246 million sounds like a lot until you stack it against what American companies are pulling in. Figure AI is at $39 billion valuation. Physical Intelligence just doubled to $11 billion.
• 1:31
  [Guest] The per-company capital in China looks modest by Silicon Valley standards, but the Forbes piece highlights ten startups in total. The broader ecosystem counts over 160 humanoid companies in China. The strategy isn't one mega-bet. It's a swarm. Dozens of companies, each adequately funded, competing on price and iteration speed while sharing a common manufacturing base.
• 1:51
  [Host] Who's actually backing these companies? Is this mostly state money or genuine private venture capital?
• 1:56
  [Guest] It's a blend that's hard to disentangle. Fourier's backers include HongShan Capital, which is the rebranded China arm of Sequoia, alongside state-adjacent funds like CICC Capital. Then Prosperity7 is Saudi Aramco's venture arm. SoftBank brings the Japanese megafund angle.
• 2:11
  [Host] Why would a Saudi oil fund back Chinese robots?
• 2:14
  [Guest] Riyadh is spending hundreds of billions diversifying away from petroleum. If humanoid robots become the labor force of the 2030s, controlling a piece of that supply chain is a hedge against oil becoming irrelevant. They're buying optionality on the post-oil economy, and they're buying it through Beijing rather than San Francisco.
• 2:32
  [Host] There's a second piece on unscarcity dot ai, "China's AI Strategy: Rival Architect, Not Villain," that makes an argument relevant here. It says Beijing isn't trying to block abundance. It's building a parallel version where Chinese institutions set the standards and control the infrastructure. The investor map you just described, Chinese, Saudi, Japanese capital all routing through Shanghai, that looks like the parallel architecture in action.
• 2:56
  [Guest] It does. And the article makes a specific distinction worth flagging. It argues the Western framing of China as a spoiler misses the point. Beijing wants to be an architect of abundance on its own terms, not a saboteur of someone else's. Fourier's coalition of backers is evidence for that thesis.
• 3:14
  [Host] What does Fourier actually build right now? Shipping or still demo phase?
• 3:19
  [Guest] Their GR-1 humanoid has been in limited deployment in rehabilitation centers and research institutions. Hospitals in China need physical therapy assistants, elder care support, patient mobility aids.
• 3:28
  [Host] So they have paying customers today, not just prototypes?
• 3:32
  [Guest] They do. And that means they're collecting real-world deployment data every day. Every patient interaction teaches the system about human movement, safety margins, operating in unpredictable environments. By the time Fourier's robots graduate to broader applications, they'll have years of clinical data that a company starting from factory floors won't have.
• 3:50
  [Host] While American startups chase the trillion-dollar general-purpose dream, Fourier is selling into hospitals and learning from every interaction. That Morgan Stanley forecast of 28,000 Chinese units this year starts to make more sense when you see companies with actual revenue, not just pitch decks.
• 4:06
  [Guest] And it connects to the 50,000-unit global projection for 2026 in that research piece. If China captures more than half that market in year one, the manufacturing learning curve tilts their direction fast.
• 4:18
  [Host] What about export controls on advanced chips? Does that slow any of this down?
• 4:23
  [Guest] It creates friction, not a wall. Fourier's exoskeleton-to-humanoid path is largely about mechanical engineering and sensor integration. The AI inference chips that run on these robots don't need bleeding-edge nodes.
• 4:33
  [Host] A humanoid lifting a patient doesn't need the same compute as training GPT-5.
• 4:38
  [Guest] Right. Where it gets complicated is the intelligence layer. If Chinese robot makers want to license cutting-edge foundation models from Google DeepMind or Physical Intelligence, sanctions could block that. But Chinese AI labs are building their own robotics models. The ecosystem is more self-sufficient than Western observers tend to assume.
• 4:56
  [Host] What worries me is supply chain concentration. If China dominates humanoid manufacturing the way it dominates solar panels and batteries, every country that imports these robots imports a dependency.
• 5:07
  [Guest] That's the less-discussed implication of this Forbes story. Individual funding rounds are news. The pattern they form is strategy. Adding humanoid robots to China's manufacturing portfolio means the physical labor infrastructure of the future could run through Chinese factories, Chinese components, Chinese standards.
• 5:22
  [Host] Could Western companies close that gap, or is it already too wide?
• 5:26
  [Guest] Tesla's building its Optimus factory at Giga Texas, Figure AI has serious manufacturing ambitions. But Chinese companies have structural advantages in component costs and government coordination. The question is whether Western buyers will pay a premium for supply chain independence the way some governments now pay more for domestically sourced semiconductors.
• 5:45
  [Host] So we're watching two competing blueprints take shape. The American model concentrates capital in a few companies betting on general-purpose intelligence. The Chinese model spreads capital across 160 companies betting on manufacturing scale and immediate deployment.
• 5:58
  [Guest] And the Fourier round crystallizes the difference. $246 million doesn't buy you a moonshot. It buys you a company already selling robots into hospitals, backed by a geopolitical coalition, positioned to ride a manufacturing wave that Morgan Stanley sees doubling this year alone.
• 6:13
  [Host] Patrick Day-glon writes in Unscarcity about how the Body, meaning robotics, is one of three converging forces reshaping civilization. Fourier's funding round is a reminder that the race to own that Body is geopolitical, and the map is being drawn right now. The book is on Amazon, audiobook on Spotify. For the research articles and the forum, head to unscarcity dot ai. This has been Minds, Bodies, and Terawatts.