The robots have indeed made significant progress compared to 2025 this year.

Yushu Robotics sold 5,500 units in a year, and this number has kept me pondering for several days.
I went to check who is buying them, and the result was a uniform group of research institutions and technology companies. They are not buying them for novelty; they are seriously engaged in research and development. To be honest, this makes me feel more at ease—this path of robotics is finally not just relying on gimmicks.
Yushu indeed has some skills; they have tackled hard nuts like motors, joints, and balance systems. Researchers can write code and run algorithms as soon as they get their hands on it, without starting from scratch with the hardware. Prices of tens of thousands to over a hundred thousand may seem expensive to ordinary people, but for research and development teams, this price for such a complete platform is already very cost-effective.
However, what I want to talk about is another matter.
Many people worry that the robotics industry will be monopolized by a certain company; I think the opposite is true. Look at the mobile phone industry; after Android became open source, it blossomed with various companies having their own ways of doing things. Robotics follows the same principle. Yushu has solidified the underlying hardware platform, which is actually paving the way for the entire industry—research teams do not need to reinvent the wheel; they can directly innovate at a higher level.
Some are working on visual recognition, some specialize in voice interaction, and others are creating applications for specific scenarios. In a few years, robots for education may look one way, robots for household chores another, and robots for companionship completely differently. It won't be dominated by one company but rather showcase diverse capabilities.
Currently, it is normal for ordinary people to buy little. Most people can't code, and when they buy one, they can only make it dance. But this situation won't last long.
From two hundred thousand to a few tens of thousands, it only takes a few years. By then, it won't just be the patent of research teams; ordinary families will also be able to have one. When that day comes, you will find that there are as many robotics companies as there are mobile phone companies now, some focusing on cost performance, some on high-end customization, and others specializing in certain niche scenarios.
The blooming landscape may arrive faster than we imagine.