Hope they make the software foss
chances are they’ll use Qwen model that Alibaba released in the open https://qwen.ai/blog?id=qwen-robotsuite
If I don’t trust a Chinese wifi-controllable power outlet, I’m probably not going to trust the Chinese humanoid AI robot.
don’t trust a
Chineseany wifi-controllable power outletnot going to trust the
Chineseany humanoid AI robotYeah; the old reason I’m naming China here is because it’s a Chinese robot.
Imagine instead of having internet companies that spy on your online activities, TVs that spies on what contents you’re watching and what games you’re playing and phones that track where you go and listen to what you say and who you are talking to, … you get one single robot that does all that for you and more
I’m just not going to trust a humanoid robot from any god damn country.
Weirdly, there probably are some countries I’d trust to build a humanoid robot. But I can’t think of any companies that I’d trust.
Call me a hater but I doubt it.
China is crazy good with technology, I expect them to become competitive in GPUs late 2027 and in memory by maybe 2028-2029.
But these are things that are well established, from production of the parts to logistics, to usage.
Humanoid robots in 2026 is like flying cars in 1950. Maybe we’ll have something like that in the future but this is way more fantastical than convenient in any way shape or form, I can’t see any scenario right now where there’s more than 10 very rich people who get to test it, take it home, and never use it again.
You can already get a UniTree humanoid for around 5k or so. The limiting factor right now is the AI system onboard rather than the hardware being hard to produce or being too expensive. And the software side is advancing at a dramatic pace. Just look at the robot performance in 2025 and 2026. It’s absolutely plausible to me that we’ll have affordable robots that can do useful work around the house within a decade.
That’s fair but what can they do? How good is their fine motor control? How good is their computer vision? How often do they need maintenance? What’s the battery life?
Traditionally these robots have a hard time with household chores because it turns out the basics are still lacking.
Don’t get me wrong, if they’re genuinely good enough for it that’s awesome, but for the average consumer it’s been “almost ready” for the last 20-30y+, so I am skeptical.
The progress with these things isn’t linear, and we can see how well these types of robots handle tasks using their hands here https://www.bastillepost.com/global/article/6361563-embodied-ai-robots-take-over-night-shift-parcel-sorting-in-chinas-tech-hub-shenzhen
So, they’re already starting to be applied in the industry now. And as with any tech, it tends to improve rapidly once it moves from the lab into actual usage. If such a robot was good enough to do common chores like cleaning up the house, doing laundry, vacuuming, etc., that would be good enough even if it doesn’t do them fast.






