Daft Punk 只说对了一半。机器人不仅会“摇滚”，还能吸引海量投资。那对以戴着机器人头盔闻名的法国音乐组合已于五年前宣布退休，但硅谷对机器人技术的痴迷正达到白热化程度。2026 年上半年，全球机器人和实体 AI 公司融资额达 334 亿美元。对于一个 2020 年全年融资额甚至未达 50 亿美元的行业来说，这是惊人的增长。
Daft Punk only had it half right. Robots don't just rock; they also draw a ton of investment. The French musicians who famously donned robot helmets retired five years ago, but Silicon Valley's obsession with robotics is reaching a fever pitch. Robotics and physical AI companies raised $33.4 billion globally in the first half of 2026. It's a stunning rise for a sector that didn't even hit $5 billion in funding in all of 2020.
我们的驻站机器人专家 Rya Jetha 撰文分析了硅谷为何押注数十亿美元赌机器人革命。这是我们新系列“机器人经济学”的首篇报道，该系列将探讨机器人的崛起及其对商业、就业和日常生活的影响。机器人技术可触达的市场规模是一大吸引力。对体力劳动的需求极其庞大——从建筑工地到家庭管家——可能性几乎无穷无尽。
Rya Jetha, our resident robotics expert, wrote about why Silicon Valley is betting billions on a robot revolution. It's the first installment in Robonomics, our new series examining the rise of robots and their impact on business, jobs, and everyday life. The size of the market that robotics can address is a big draw. The need for physical labor is so vast — from construction sites to at-home butlers — the possibilities are almost endless.
英伟达 CEO 黄仁勋是实体 AI 的坚定支持者，他将该行业估值定在……50 万亿美元。机会摆在那里，并不意味着容易变现。以下是阻碍机器人革命的一些最大痛点：
Nvidia CEO Jensen Huang, a huge proponent of physical AI, valued the industry at … $50 trillion. Just because the opportunity is there doesn't mean it'll be easy to realize. Here are some of the biggest pain points standing in the way of the robotics revolution:Not an easy hand-off: When talking about robots, the discussion typically starts and ends with human hands.
交接不易：谈到机器人，讨论通常始于人手，也终于人手。原来我们的双手相当特别，很难复刻。诚然，并非每家公司都想复制人手。但许多任务需要我们拥有的灵巧度。如何将这些能力植入机器人，一直难以攻克。
It turns out our paws are pretty special, and quite difficult to recreate. Granted, not every company wants to replicate the human hand. But a lot of tasks require the dexterity we have. Figuring out how to build those capabilities into robots hasn't been easy. Hardware is hard: Building robots has gotten considerably cheaper than years ago thanks to advancements in computing and hardware.
硬件难搞：得益于算力和硬件的进步，造机器人比几年前便宜得多。但它仍比大多数软件开发昂贵得多。重写几行代码是一回事，打造全新原型又是另一回事。随着你试图扩大实体生产，这些问题会被放大。供应链上微小的失误若传导太远，就会变成大麻烦。就问问埃隆·马斯克吧。
It's still a much more expensive endeavor than most software work, though. It's one thing to rewrite some code. It's another to build an entirely new prototype. Those problems get magnified as you try to scale up physical production. The smallest mistake becomes a bigger problem if it gets too far down your supply chain. Just ask Elon Musk about it.
企业常将 AI 工具宣传为增强而非取代人类工人的手段。这究竟是否属实，目前仍有争议。制造能完成与人类相同体力劳动的机器人，留给“另一种解读”的空间要小得多。仍有人不认为这是零和博弈。机器人初创公司 Machina Labs 的首席执行官爱德华·梅尔此前曾告诉我，他的新工厂雇佣的人类工人数量，与一家没有机器人的工厂持平。
Human pushback: Companies often pitch AI tools as a way to enhance, not eliminate, human workers. Whether that's the reality is up for debate. Building robots to perform the same physical labor as humans leaves less room for interpretation. Some still don't see it as a zero-sum game. Edward Mehr, the CEO of robotics startup Machina Labs, previously told me his new factory had the same number of human workers as one at a robot-free facility.
如果机器人真开始取代人类，它们还得应对蓝领领域特有的另一个微妙因素：工会。AI 已让部分蓝领工人受益，其中包括参与建设 AI 数据中心的工人。这种建设热潮让技术工种大获其利，甚至一些全国性劳工工会都站在了开发商一边。但如果机器人意味着取代他们的成员，这些工会态度就会变得不那么友善了。
If robots do start replacing humans, they'll have to navigate another nuance specific to the blue-collar world: unions. AI has benefited some blue-collar workers, including those involved in building AI data centers. That construction has been a boon for tradespeople, with some national labor unions even siding with developers. They'll be a lot less friendly if robots mean replacing their members, though.