On August 19, Unitree Technology made its market debut, with shares skyrocketing by over 600% at the open and briefly hitting a high of 1,100 yuan. While humanoid robots have yet to enter households on a massive scale, the capital markets have already rushed to stake their claim. But why?

If one only looks at the company's current revenue and profits, such a lofty valuation is difficult to justify. What the market is genuinely paying for is a much grander narrative: the robots of the future may not be sold as machines, but rather as a subscription to "labor." This is precisely where the most compelling imagination lies for humanoid robotics.

In recent years, the most viral robot videos often feature running, dancing, backflips, and boxing. Netizens watch for the spectacle, but companies are crunching the numbers. A factory won't shell out hundreds of thousands of yuan for a robot just because it can perform a backflip. What business owners truly care about is: How many hours a day can it work? Can it handle moving goods, tightening screws, or performing inspections? How long does it take to repair when it breaks down? And how many years will it take to recoup the investment?

At its core, the biggest hurdle for robots is no longer "looking like a human," but "working like one." Once this threshold is crossed, the business model transforms completely. Previously, selling robots was essentially selling equipment—a one-for-one transaction. In the future, if robots can continuously learn and adapt to various roles, what gets sold could be a scalable, replicable "machine labor force." At that point, companies like Unitree would not just be competing in the robotics market, but for a far larger labor market underpinning manufacturing, logistics, commercial services, and even home care.

This is why capital is willing to bet early. China, in particular, holds a significant advantage—an abundance of application scenarios. Vast manufacturing floors, warehouses, ports, and shopping malls provide a natural "training ground" for robots. The more tasks robots perform, the more data they accumulate; the more data they have, the smarter they become; and the smarter they get, the more jobs they can take on. This is what could build a true moat.

Of course, a 600% surge on the first trading day in no way signifies that Unitree has already won. On the contrary, the faster the stock price runs, the harder the performance must chase. Capital may pay in advance for the robot world of a decade from now, but the company must ultimately deliver on its story through orders, profits, and real-world applications.

So, when evaluating robotics companies going forward, it's wise to watch fewer backflips and instead ask three critical questions: How much work can it do? How much money can it save? And how soon will it pay for itself? When these numbers finally add up, humanoid robots will cease to be just the protagonists of tech news and will become ordinary employees on the assembly line. Unitree's listing is merely capital pressing the accelerator first; what truly decides this race is when robots begin to earn money for people.