- Unitree G1 demonstrates endurance, acrobatics, and an 'Anti-Gravity' mode in controlled testing.
- China accelerates civilian and industrial deployment and analyzes military applications, amid internal ethical debate.
- Million-dollar investments contrast with Rodney Brooks' technical skepticism about feasibility and safety.
- The market is moving forward with pilots; thousands of units are expected in the short term and growth is expected for decades to come.
At the same time, the conversation opens up to broader issues: investment, civil and military uses, and the limits of safety and ethics. This cocktail, which mixes technical prowess with expert warnings, leaves us with a panorama in which the future of humanoids It feels close, but with many pieces still to fit together.
A technical leap: the Unitree G1 and its "Antigravity" mode
Unitree's latest video shows a "violence test" where several operators kick and push the G1 without hesitation; despite the impacts, the robot maintains its balance, wobbles just enough, and recovers posture with surprising speedThe firm claims that the speed of the footage is not altered and that the G1 stands up in about one or two seconds.
Beyond the resistance, they are appreciated acrobatics, mid-air spins and martial-inspired movements that test their dynamic control. The test platform seen in the video is based on the Unitree G1, with 23 degrees of freedom, about 35 kilos in weight and 1,32 meters in height, a configuration that allows you to execute quick steps, climb uneven surfaces and manipulate objects with three-fingered hands.
The key to control is in perception: the system combines depth cameras and LiDAR sensors to understand the environment and plan movementsThe company itself emphasizes that these are demonstrations in a controlled environment and should not be replicated; it's no small feat to coordinate such energy and precision without putting those around you at risk.
On a practical level, this unit is geared toward research and development and has also proven itself in domestic "butler" type functions. There are even devices that, for less than $6.000, They perform somersaults and other advanced mobility exercises, a sign that the technical barrier to entry is beginning to lower., although with nuances regarding reliability, cost and real usefulness.
China accelerates: from factories to military ambitions
On the geopolitical and industrial front, China is preparing the ground for an ambitious deployment starting in 2025, with the idea of introducing more than a million AI robots in areas ranging from the home and logistics to hospitals and production plants. This push is note already in pilot projects and in technical exhibition events.
The military dimension is also not left out. An article in the People's Liberation Army Daily points out that humanoids could combine "tactical flexibility and strategic deterrence" in smart warfare scenarios, thanks to their bionic structure and ability to adapt to complex environments like ruins or stairs.
Examples such as the Tien Kung Ultra, a 1,8-metre humanoid that won the first half marathon for robots held in Beijing, are cited as proof that the bipedal design facilitates navigation, something relevant when thinking about rescue operations or conflict theaters.
Chinese strategists imagine mixed formations with soldiers, humanoids and autonomous teams: robots that provide support in mobility and shooting, or even act as decoys to divert enemy fireThe so-called "uncanny valley" is also mentioned, a possible psychological deterrent effect when confronted with near-human machines.
Ethics, law, and the specter of killer robots
Not everyone shares this enthusiasm. In the same EPL Journal, three authors —Yuan Yi, Ma Ye and Yue Shiguang—warned that normalizing the use of humanoids in combat could lead to accidental deaths and indiscriminate killings, with legal consequences and moral reproach that are difficult to manage.
His proposal is to strengthen ethical and legal research, and regulate and restrict the use of humanoids in war scenarios while the technology matures. They emphasize that understanding of these platforms remains "in its infancy," with insufficient exploration and the risk of loss of control if too much is done.
Investments, expectations and critical voices
At the same time, private financing continues unabated. Startups like figure or Apptronik have attracted hundreds of millions of dollars and interests from big tech companies, with collaborations aimed at combining next-generation hardware and advanced artificial intelligence. The The dominant narrative promises humanoids in factories, offices, and eventually homes..
But one of the most respected names in robotics, Rodney Brooks (co-founder of iRobot and former professor at MIT), urges caution. In his opinion, it is "fantasy" to think that robots will learn manual dexterity just by watching videos: the human hand concentrates around 17.000 touch receptors and there is no equivalent database that allows replicating that fineness in the short term.
Brooks also warns of a little-discussed angle: physical securityStaying upright requires a lot of energy, and when a humanoid falls, that energy is released all at once. Doubling the size, he calculates, multiplies the potentially damaging energy in a fall by eight; by pure physics, A large robot can be dangerous in uncontrolled environments.
For him, the most useful way in the medium term is through machines with wheels, multiple arms, and specialized sensors, designed for specific tasks and not to look like us. That is, Fewer "metal people" and more effective, efficient, and safe tools.
A recent study by the METR organization adds nuances to expectations about AI: although Developers believed they were working 20% faster when using these tools, but ended up being 19% slower in controlled tests.Data like this helps to adjust optimism to operational reality.
When will they arrive home? A market that is still taking off
On the commercial front, actual deployments remain limited and highly piloted. Bank of America estimates that about 18.000 humanoid units this year already a growth path that, in ambitious scenarios, could approach one billion robots by 2050. These are figures for decades to come, not for immediate maturity.
China, meanwhile, is investing heavily to build a market that doesn't yet exist, trusting that the AI will solve autonomy, reliability and utility as it matures. Today, this promise coexists with technical limitations and operating costs that encourage us to move forward carefully.
In the field of public visibility, events such as the recent ones are held Humanoid Robot Olympics in China, with industrial and sports exhibitions. Beijing has elevated humanoid robotics to a state priority with its sights set on leading "the next technological revolution" after mobile phones and electric vehicles.
The picture that emerges combines notable advances (case G1), a significant state and industrial commitment, and a technical-ethical debate that cannot be ignored; with these elements, humanoids gain traction, but its mass adoption will depend on how safety, dexterity and real value are resolved in everyday uses.
I am a technology enthusiast who has turned his "geek" interests into a profession. I have spent more than 10 years of my life using cutting-edge technology and tinkering with all kinds of programs out of pure curiosity. Now I have specialized in computer technology and video games. This is because for more than 5 years I have been writing for various websites on technology and video games, creating articles that seek to give you the information you need in a language that is understandable to everyone.
If you have any questions, my knowledge ranges from everything related to the Windows operating system as well as Android for mobile phones. And my commitment is to you, I am always willing to spend a few minutes and help you resolve any questions you may have in this internet world.