U.S. tech giants are betting big on humanoid robots — but China's already ahead, analysts say (1 Viewer)

[slavablueberryjam]

https://www.cnbc.com/2025/03/28/china-already-ahead-of-us-in-humanoid-robot-race-analysts-say-.html

The best way to counter this in America is to cut education funds.

 

[HonestJoe]

Something of an aside, but I've never understood the big push for humanoid robots in the first place. It is extremely difficult to make a robot that can reliably balance on two legs and so outside of some entertainment fields, I would suggest that different form factors would be much more suited to the vast majority of uses. We already use a vast range of robotic devices today and none of them are fully humanoid because that would make zero sense. In South Africa, traffic lights are called "robots" (and they're really not wrong), yet they literally replaced a human being standing in the middle of junctions waving their hands.

 

[dmpi]

China don't need robots because Chinese human labor is already very cheap and getting even cheaper.

 

[Gatsby]

Something of an aside, but I've never understood the big push for humanoid robots in the first place. It is extremely difficult to make a robot that can reliably balance on two legs and so outside of some entertainment fields, I would suggest that different form factors would be much more suited to the vast majority of uses. We already use a vast range of robotic devices today and none of them are fully humanoid because that would make zero sense. In South Africa, traffic lights are called "robots" (and they're really not wrong), yet they literally replaced a human being standing in the middle of junctions waving their hands.

Humanoid robots are important since most of our machines, tools, and living spaces are designed for human bodies. If we wanted to make, for example, a housekeeping robot, a humanoid would be the best choice since they'd have to work appliance buttons with fingers and navigate human-sized doorways and stairs. Maybe over time that will change, as we start designing our spaces with robots in mind.

 

[HonestJoe]

Humanoid robots are important since most of our machines, tools, and living spaces are designed for human bodies. If we wanted to make, for example, a housekeeping robot, a humanoid would be the best choice since they'd have to work appliance buttons with fingers and navigate human-sized doorways and stairs. Maybe over time that will change, as we start designing our spaces with robots in mind.

That assumes such general purpose machines are desired, or truly practical. I wouldn't expect them to have to work appliances, they would be the appliances.

Think of things like self-service checkouts in supermarkets, ordering consoles in fast-food restaurants or self-driving cars. They didn't create machines to simply replicate the human staff and have them use the same equipment, they redesigned the systems with the automation in mind. I don't see any reason for future steps along the same path to significantly change that principle.

 

[The Mark]

China don't need robots because Chinese human labor is already very cheap and getting even cheaper.

No, it isn't.

In fact I've encountered multiple mentions over the past 5-10 years that labor cost in China is rising, which is why some production was moved to other nations.

Edit: And that's not even considering whether China itself wants to remain with very cheap labor, at the cost of living standards.

 

[dmpi]

No, it isn't.

In fact I've encountered multiple mentions over the past 5-10 years that labor cost in China is rising, which is why some production was moved to other nations.

Edit: And that's not even considering whether China itself wants to remain with very cheap labor, at the cost of living standards.

China moved production to Vietnam, Malaysia, Thailand and Cambodia to get around Trump's first term tariffs. China's in a depression, they're suffering from deflation, they're housing is in free-fall, young people cannot find a job. You better believe that their labor is cheap & getting cheaper.

 

[Airyaman]

America got cocky. We thought we were thought leaders. For years, it was "The US creates, the world strives to improve".

You can't stay number #1 unless you're always moving forward. We aren't anymore.

 

[Hamish Howl]

Humanoid robots are dumb as hell.

 

[Lord Tammerlain]

China moved production to Vietnam, Malaysia, Thailand and Cambodia to get around Trump's first term tariffs. China's in a depression, they're suffering from deflation, they're housing is in free-fall, young people cannot find a job. You better believe that their labor is cheap & getting cheaper.

Chinese skilled labor is more expensive than India, most of South East Asia, Mexico etc

 

[Lord Tammerlain]

Something of an aside, but I've never understood the big push for humanoid robots in the first place. It is extremely difficult to make a robot that can reliably balance on two legs and so outside of some entertainment fields, I would suggest that different form factors would be much more suited to the vast majority of uses. We already use a vast range of robotic devices today and none of them are fully humanoid because that would make zero sense. In South Africa, traffic lights are called "robots" (and they're really not wrong), yet they literally replaced a human being standing in the middle of junctions waving their hands.

The balance issue seems to be generally solved


The plant I work at, could have 80% of the people on the floor replaced by humanoid robots with good vision systems and programming. We are talking basic labor functions along a production line. Other industrial robots are far more limited, and are used as well but a humanoid robot could do quite a few more functions

 

[BirdinHand]

China don't need robots because Chinese human labor is already very cheap and getting even cheaper.

If that’s true, then China is even further ahead of us.

Their GINI coefficients for wealth disparity, etc are closing gaps that are only getting worse here

 

[slavablueberryjam]

China moved production to Vietnam, Malaysia, Thailand and Cambodia to get around Trump's first term tariffs. China's in a depression, they're suffering from deflation, they're housing is in free-fall, young people cannot find a job. You better believe that their labor is cheap & getting cheaper.

China has more skilled labor when it comes to trades, etc. They are meant to be the manufacturing hub of the world. That's what they've been working on. Honestly, Americans are not understanding of what Trump is doing because they bought into the whole 'biden' economics ruined America-and now Trump will fix that. Trump won't fix it. Everything points to Trump-based on testimonies from those around him-is that Trump...is an actual idiot. So, yeah, you get what you vote for.

 

[HonestJoe]

The balance issue seems to be generally solved

I never said it hadn't, only that it is difficult to do. Why invest in all of the hardware and software necessary for a machine to balance on two legs when there are a range of much easier (and thus cheaper) options depending on the requirements.

The plant I work at, could have 80% of the people on the floor replaced by humanoid robots with good vision systems and programming. We are talking basic labor functions along a production line. Other industrial robots are far more limited, and are used as well but a humanoid robot could do quite a few more functions

But why would you need complex multi-purpose machines for specific tasks in a fixed environment at all? There could certainly be use cases for more flexible machines to the current industrial robots, but even then, the best solution is never going to be a human shape and size, where half of it's capacity is going to be used simply keeping it upright.

Humanoid robots are like jet-packs or flying cars. They all looked cool in the classic science fiction of our youth, but while it is possible to make them, and a few people do, they'll never be practical for the majority of general purpose use cases.

 

[L88]

https://www.cnbc.com/2025/03/28/china-already-ahead-of-us-in-humanoid-robot-race-analysts-say-.html

The best way to counter this in America is to cut education funds.

Well, sure, they're way ahead. They have all their research and all of Western world's research collected and being used by their US college-educated engineers and scientists.

 

[Lord Tammerlain]

to

I never said it hadn't, only that it is difficult to do. Why invest in all of the hardware and software necessary for a machine to balance on two legs when there are a range of much easier (and thus cheaper) options depending on the requirements.

But why would you need complex multi-purpose machines for specific tasks in a fixed environment at all? There could certainly be use cases for more flexible machines to the current industrial robots, but even then, the best solution is never going to be a human shape and size, where half of it's capacity is going to be used simply keeping it upright.

Humanoid robots are like jet-packs or flying cars. They all looked cool in the classic science fiction of our youth, but while it is possible to make them, and a few people do, they'll never be practical for the majority of general purpose use cases.

We do a wide variety of different packages, different jug size different cap size different number of jugs in a box. Traditional industrial robots that we currently use need different end of tool attachments, that a humanoid robot with " hands would not. The humanoid robots offer more flexibility which in the case of the company I am at requires ( smaller mid size with revenues between 60 to 80 million per year). Combine that with a significant lack of industrial robot technicians in western Canada, a humanoid robot with lots of flexibility in what they can do is ideal.

A Dobot we use for putting empty jugs on a packaging line costs approximately $15000, a unitree humanoid robot would approximately be the same price in a few years, while offering a much larger scope in what it could do.

For large scale manufacturing with very similar products being run day in day out, human robots would not be a great choice

 

[Infinite Chaos]

At one point in the video - it talks about robot vs robot livestream boxing matches in the next 30 days.

I find the prospect as interesting and exciting as the idea once mooted of robot controlled Formula 1 racing which is to say - not at all my "cup of tea."

That said, the capabilities are remarkable compared to even 10 years ago - I just wonder though if human labour and intelligence is supplanted by robotic and AI - who will there be with any earnings to buy the products and services produced by this new technology when we are all replaced at the workplace?

 

[Lord Tammerlain]

At one point in the video - it talks about robot vs robot livestream boxing matches in the next 30 days.

I find the prospect as interesting and exciting as the idea once mooted of robot controlled Formula 1 racing which is to say - not at all my "cup of tea."

That said, the capabilities are remarkable compared to even 10 years ago - I just wonder though if human labour and intelligence is supplanted by robotic and AI - who will there be with any earnings to buy the products and services produced by this new technology when we are all replaced at the workplace?

The boxing match is of course a novelty event.

The main point especially if autonomous is that they are capable of it, and getting back up.

If programming them to do something is as simple as showing them what is acceptable or not, vs writing pages of code, then low skill labor is dead, AI of course is coming for white collar positions as well

 

[HonestJoe]

We do a wide variety of different packages, different jug size different cap size different number of jugs in a box. Traditional industrial robots that we currently use need different end of tool attachments, that a humanoid robot with " hands would not.

That would just involve adjustable rather than switchable attachments though. There are probably reasons that hasn't already been done (cost of the additional complexity doesn't balance the time lost manually switching attachments) and even if those adjustable attachments were more "hand-like", that wouldn't require the rest of the robot to be fully humanoid (or fully automated).

Combine that with a significant lack of industrial robot technicians in western Canada, a humanoid robot with lots of flexibility in what they can do is ideal.

If you have a shortage of technicians there, I'm not convinced bringing in much more complex, cutting edge machinery would be a good solution. The more a machine can do, the more there is to go wrong.

 

[Infinite Chaos]

The boxing match is of course a novelty event.

The main point especially if autonomous is that they are capable of it, and getting back up.

Of course, my take was and is on the prospect of autonomous robot vs autonomous robot in sport.

 

[Girlpower]

Something of an aside, but I've never understood the big push for humanoid robots in the first place. It is extremely difficult to make a robot that can reliably balance on two legs and so outside of some entertainment fields, I would suggest that different form factors would be much more suited to the vast majority of uses. We already use a vast range of robotic devices today and none of them are fully humanoid because that would make zero sense. In South Africa, traffic lights are called "robots" (and they're really not wrong), yet they literally replaced a human being standing in the middle of junctions waving their hands.

The reason for humanoid form is so they can traverse human areas which are generally constructed with humanoids in mind.

Stairs, doorways, vehicles......sure you could use other forms but they may not be as effective.

Specifically we are talking about robotic soldiers.

Whoever designs one that meets criteria locks in huge military contracts.

 

[HonestJoe]

The reason for humanoid form is so they can traverse human areas which are generally constructed with humanoids in mind.

But the vast majority of use-cases don't require that, and even where they do, a fully humanoid form factor isn't automatically the best solution. We already have a massive range of "robots" in all sorts of contexts and the majority of them don't need to move around at all. Those which do typically use different forms, like wheels or flying, because they're much more efficient ways to manage the same requirements.

Specifically we are talking about robotic soldiers.

Again, there are already "robots" used in military contexts, but they are generally either flying drones of various forms and sizes, wheeled/tracked or occasionally four-legged. There is nothing special about a six foot, two arm, two legs and a head that makes it the default solution to any set of requirements.

 

[ataraxia]

I just wonder though if human labour and intelligence is supplanted by robotic and AI - who will there be with any earnings to buy the products and services produced by this new technology when we are all replaced at the workplace?

That might be a good time to start talking about Universal basic income (UBI).

 

[Infinite Chaos]

That might be a good time to start talking about Universal basic income (UBI).

In the Star Trek world, people don't have regular jobs and spend their time on leisure / exploring the universe etc but I think our AI and Robot future will be more akin to Soylent Green.

 

[ataraxia]

In the Star Trek world, people don't have regular jobs and spend their time on leisure / exploring the universe etc but I think our AI and Robot future will be more akin to Soylent Green.

Why that and not Star Trek? Is it that you’re just so disillusioned with our society today?