Join the movement and participate on Uniswap.
Participate
Walk into many robotics labs in China and you’ll spot some familiar patterns: programmable robots, open APIs, companies open to collaborate, and distributors lining up to take new units into warehouses, research labs, and universities.
By contrast, walk into many robotics labs in Silicon Valley, and you’ll see the opposite: closed systems, locked-down software, no APIs, no distributors and founders insisting that no third-party code will ever run on their robots.
There are, naturally, many labs that do not conform to this oversimplification, but the biggest and most funded robotics companies in the West in particular do overwhelmingly take this closed approach.
Do these two fundamentally different approaches represent two different visions of how the global robotics market will unfold or are they being shaped by the economic realities of the markets that fund them?
We think it’s the latter: funding for robotics in the US is pushing US robotics companies in a direction that is leading to successful fundraising, while at the same time leaving the overall market at a technological and competitive disadvantage.
What follows is a generalization about US and Chinese robotics companies and their approaches based on our interactions with them. While there are outliers that require more nuance, we think this generalization still merits discussion.
China’s robotics industry is scaling at a pace the West still struggles to come to terms with. The country deployed more robots last year than the rest of the world combined. In addition to manufacturing the most robots, China was the biggest market for them.
China’s humanoid ecosystem alone spans dozens of companies: covering the full range of humanoids from Unitree and Booster to AgiBot and Galbot. Many of these companies are not familiar names in the West, even to investors and industry professionals.
What unites them?
Chinese robots are:
- Programmable and often open source
- Sold primarily to researchers, integrators, and developers
- Distributed through partners, not directly by the manufacturers
- Shipped in comparatively large quantities, even if the software isn’t “finished”
The Chinese strategy acknowledges a simple truth: robots don’t become useful in the real world until someone teaches them to do real-world tasks. And these companies are happy to outsource that teaching to integrators, researchers, and ecosystems.
This approach has created:
- A proliferation of robotics companies
- The world’s largest robotics research base
- The world’s most affordable humanoids and quadrupeds
- The deepest supply chain and fastest iteration speed
It’s messy. It’s distributed. But it works and it's scaling. China is shipping robots: if you go to an academic institution in the West they are most likely conducting their research on a Unitree.
In the US, by contrast, the largest robotics companies overwhelmingly refuse to allow:
- Third-party code
- External integrators
- Open APIs
- Collaborative ecosystems
Instead, they appear to share a common belief or perhaps a shared pitch deck:
“We are building a general-purpose humanoid with human-level intelligence that will work on day one with zero integration.”
American founders know that if they admit they’re building “just another programmable robot” they face an obvious challenge: China will out-scale them, out-price them and out-ship them.
So the pitch has to be something along the lines of:
“Programmable robots are obsolete. We’re skipping straight to AGI.”
It’s also explains why these robots:
- Are closed ecosystems
- Are mostly not shipping and won’t ship until “the AI is ready”
- Are not sold and deployed by distributors and integrators
- Are pitched as labor-as-a-service with hourly contracts
It’s a winner-takes-all philosophy, heavily shaped by American venture capital.
But it comes with a cost.
In this context, it's a worrying development that K-Scale, one of America’s notable open-source humanoid startups, is winding down.
In a message to their community, the founder openly admitted:
- They couldn’t find a lead investor
- They couldn’t fund the tooling for mass production
- They couldn’t match the economies of scale achieved in China
K-Scale’s failure is symbolic of the West’s structural misunderstanding of the robotics market.
Open-source platforms are crucial to building robotics talent. If the US doesn’t have any, then American students will learn on Chinese robots, American labs will research on Chinese robots and American developers will build for Chinese ecosystems.
Put bluntly: American investors don’t understand robotics, and it shows.
When Silicon Valley VCs think “robotics,” they think: Tesla, Figure and 1X. That’s it. Maybe Unitree, if they’re unusually informed.
But they rarely grasp what's happening across Shenzhen, Hangzhou, or Dongguan.
The result?
A tiny cluster of over-funded American “moonshots” are taking on a massive, diversified, fast-moving ecosystem in China. The lack of collaboration and interest in working with integrators or distributors also kills any possibilities of a grassroots robotics community in the US, resulting in a talent and capability gap that widens every year.
The current state of play pits a sophisticated supply-chain of robotics companies that are shipping and iterating on their hardware against a few enormously funded companies with deep pockets.
One strategy has the potential to support a network of robotics manufacturers, integrators, distributors, not to mention the grassroots: enthusiasts and researchers.
The other bets everything on a few elite teams cracking AGI first and then selling labor-as-a-service.
In a broader sense, a healthy robotics market shouldn’t be winner-takes-all.
It shouldn’t concentrate all robotics literacy in one country or a small number of companies.
A healthy ecosystem needs:
- Open platforms
- Developer access
- Global collaboration
- Privacy-preserving infrastructure
If the West wants to remain competitive, it must invest not only in closed AGI moonshots but also in open, programmable, widely accessible robotics platforms.
It’s unlikely a handful of companies will be able or willing to meet the needs of every robotics customer.
Those use cases will need to be met by developers, integrators and researchers teaching robots how to work in the real world. And they need developer-friendly platforms to work on.
We think the real world web is a part of this puzzle. A shared map that enables real world apps, streamlines the deployment of integrated robots in real world locations and preserves the privacy of every device on it.
Learn more about the real world web here.
Aukiはポーズメッシュという地球上、そしてその先の1000億の人々、デバイス、AIのための分散型機械認識ネットワークを構築しています。ポーズメッシュは、機械やAIが物理的世界を理解するために使用可能な、外部的かつ協調的な空間感覚です。
私たちの使命は、人々の相互認知能力、つまり私たちが互いに、そしてAIとともに考え、経験し、問題を解決する能力を向上させることです。人間の能力を拡大させる最も良い方法は、他者と協力することです。私たちは、意識を拡張するテクノロジーを構築し、コミュニケーションの摩擦を減らし、心の橋渡しをします。
ポーズメッシュは、分散型で、ブロックチェーンベースの空間コンピューティングネットワークを動かすオープンソースのプロトコルです。
ポーズメッシュは、空間コンピューティングが協調的でプライバシーを保護する未来をもたらすよう設計されています。いかなる組織の監視能力も制限し、空間のプライベートな地図の自己所有権を奨励します。
分散化はまた、特に低レイテンシが重要な共同ARセッションにおいて、競争優位性を有します。ポスメッシュは分散化運動の次のステップであり、成長するテック大手のパワーに対抗するものです。
アウキ・ラボはポスメッシュにより、ポーズメッシュのソフトウェア・インフラの開発を託されました。
