Robot Goes Rogue in China

jujube said:
If my robot maid attacks me, is that domestic violence?
Click to expand...
Though doubt will be around to see it what a treat robotic cleaning , gardening , etc will be .

A robot or whatever they are called who does laundry , food shopping , cleaning .

Now that would be nice .

 

Pepper said:
You always give wonderful descriptions! Trust you more than China any day!
Click to expand...

Just for the sake of clarity -- the description in the original post is taken word-for-word from the Daily Mail article. It’s always best to be clear on what’s quoted and what are someone’s own words. As it avoids later confusion as a thread develops.

If you don't want to click on the article link for the reasons you have already given, I could post it here with its remaining paragraphs if you like.
 
Last edited:
" Footage making the rounds on social media shows an AI-powered humanoid robot having a startling meltdown without warning. The video, apparently leaked from a Chinese lab, has become bigger than itself, triggering debate over artificial intelligence, robotics safety, and the international race to deploy robotic humanoids at scale.As general-purpose humanoid robots become more accessible, a shift already underway, such moments are bound to become more frequent and potentially more dangerous.

The 31-second clip went viral after a reddit user posted it in the oddly terrifying sub reddit with the title “Robot on hook went berserk all of a sudden.” The community’s users leaned into science fiction tropes with jokes about the Terminator, Skynet, Robocop, and Iron Man.The clip spread across social media, eliciting a range of amusement, technical curiosity, and skepticism. The incident has sparked broader conversations about the current state of humanoid robotics, the challenges developers face, and the public’s perception of rapidly advancing artificial intelligence technologies.

The viral video reflects a divide in how China and the West approach humanoid robotics development amid an escalating technological arms race.Messaging from Western companies like Boston Dynamics, Apptronik , Agility Robotics, and even Tesla center around gradual rollout, formal safety systems, and engineering robustness. They typically avoid public demos that could do brand damage.In contrast, Chinese companies like Unitree and EngineAI are putting their humanoid robots through grueling public tests, including synchronized dance performances unpolished and even half marathons. There’s less fear of embarrassment, and more willingness to test prototypes in public view.

Footage making the rounds on social media shows an AI-powered humanoid robot having a startling meltdown without warning. The video, apparently leaked from a Chinese lab, has become bigger than itself, triggering debate over artificial intelligence, robotics safety, and the international race to deploy robotic humanoids at scale.

As general-purpose humanoid robots become more accessible, a shift already underway, such moments are bound to become more frequent and potentially more dangerous. The 31-second clip went viral after a reddit user posted it in the oddly terrifying sub reddit with the title “Robot on hook went berserk all of a sudden.” The community’s users leaned into science fiction tropes with jokes about the Terminator, Skynet, Robocop, and Iron Man.

The clip spread across social media, eliciting a range of amusement, technical curiosity, and skepticism. The incident has sparked broader conversations about the current state of humanoid robotics, the challenges developers face, and the public’s perception of rapidly advancing artificial intelligence technologies.

The viral video reflects a divide in how China and the West approach humanoid robotics development amid an escalating technological arms race. Messaging from Western companies like Boston Dynamics, Apptronik , Agility Robotics, and even Tesla center around gradual rollout, formal safety systems, and engineering robustness. They typically avoid public demos that could do brand damage.

In contrast, Chinese companies like Unitree and EngineAI are putting their humanoid robots through grueling public tests, including synchronized dance performances unpolished and even half marathons. There’s less fear of embarrassment, and more willingness to test prototypes in public view.

The Shocking Rise of Unitree Robotics, Wang Xingxing, Leaders in China's Push for AI Humanoid Robots • The Shocking Rise of Unitree Robotics... 0:00 - The Viral Robot Meltdown 0:36 -

Reddit Reacts: Sci-Fi Tropes & Memes 1:03 - Why This Footage Sparked Real Debate 1:16 - Westwood Robotics Responds 1:37 - Themis V2: Safety-First Design 2:22 - East vs. West: A Robotics Arms Race 3:08 - Unitree’s Public Test Philosophy 3:20 -

What Happened in the Bilibili Clip? 3:35 - Vibe Coding & AI Safety Risks 4:03 - Robot Kombat & Real Steel Vision 4:49 - Unitree’s Combat Bots & VR Pilots 5:05 - Inside Unitree’s Rise: H1 & G1 Overview 6:06 - The Humanoid Robot Gold Rush 6:29 -

Was the Viral Video a Leak or a Stunt? 7:01 - Sim Footage Reveals the Truth? 7:33 - A Pattern of Public Malfunctions 8:09 - The Future of Unitree & Embodied AI

 
I’ve since come across this explanation of what might have likely happened. It matches part of what I previously mentioned as a possibility. In that the robots programming code might have been reacting to it being tethered in a way it wasn't programmed to handle, after its initial movement.

Maybe not so much a robot going rogue, but more of a software control issue? From this link, Chinese robot suddenly loses control - makes strange movements, attacks a person :

"During a public demonstration, the H1 humanoid robot developed by Chinese company Unitree suddenly began to move erratically and uncontrollably, surprising those present and causing a sense of uncertainty about the durability of its technology.

What really happened? Most likely, the robot was trying to maintain its balance in unusual conditions.

This robotic model is equipped with an advanced stability maintenance system, which enables it to react in real time to sudden changes in attitude or surface.

However, during this demonstration, the robot was attached to a safety rope attached to its head — a common practice for safety reasons during public testing.

The problem is that the robot's stabilization code is not designed to work in situations where it is constrained by a fixed point on its head. The tethering limited its ability to make natural corrections to its position, leading to a series of increasingly aggressive attempts to regain balance.

Since the initial movements — small and careful — did not produce the desired effect, the robot's internal control algorithm reacted by increasing the intensity and amplitude of the movements, in an attempt to "escape" from what it interpreted as severe destabilization.

This led to increasingly erratic and dramatic behavior, to the point where the robot seemed to have completely lost control.

For observers, the moment was shocking: the robot moved in a strange, unpredictable manner, and for a few moments even created the impression of a frightening futuristic scenario — as if it had been disconnected from human control..."
 
Magna-Carta said:
I’ve since come across this explanation of what might have likely happened. It matches part of what I previously mentioned as a possibility. In that the robots programming code might have been reacting to it being tethered in a way it wasn't programmed to handle, after its initial movement.

Maybe not so much a robot going rogue, but more of a software control issue? From this link, Chinese robot suddenly loses control - makes strange movements, attacks a person :

"During a public demonstration, the H1 humanoid robot developed by Chinese company Unitree suddenly began to move erratically and uncontrollably, surprising those present and causing a sense of uncertainty about the durability of its technology.

What really happened? Most likely, the robot was trying to maintain its balance in unusual conditions.

This robotic model is equipped with an advanced stability maintenance system, which enables it to react in real time to sudden changes in attitude or surface.

However, during this demonstration, the robot was attached to a safety rope attached to its head — a common practice for safety reasons during public testing.

The problem is that the robot's stabilization code is not designed to work in situations where it is constrained by a fixed point on its head. The tethering limited its ability to make natural corrections to its position, leading to a series of increasingly aggressive attempts to regain balance.

Since the initial movements — small and careful — did not produce the desired effect, the robot's internal control algorithm reacted by increasing the intensity and amplitude of the movements, in an attempt to "escape" from what it interpreted as severe destabilization.

This led to increasingly erratic and dramatic behavior, to the point where the robot seemed to have completely lost control.

For observers, the moment was shocking: the robot moved in a strange, unpredictable manner, and for a few moments even created the impression of a frightening futuristic scenario — as if it had been disconnected from human control..."
Click to expand...
Initially it looked like the robot was trying to slap him but after you can see simply slapping wildly though would hate to be caught with one of those robot slaps .
 
Tish said:
I think it is staged, but you never know.
Click to expand...
I think it was real but again you never know .

It happened some time ago and video got out first factory denied it then admitted it happened but taken care of .

I would not mind like a robot that cleaned but would not want it to stay over night due to fear it would kill me not to mention being a single man sometimes I have brushed my teeth in kitchen sink which one of my daughters when visiting was irate about and a cleaning robot might not over look this and terminate me .
 
Magna-Carta said:
Just for the sake of clarity -- the description in the original post is taken word-for-word from the Daily Mail article. It’s always best to be clear on what’s quoted and what are someone’s own words. As it avoids later confusion as a thread develops.

If you don't want to click on the article link for the reasons you have already given, I could post it here with its remaining paragraphs if you like.
Click to expand...
Thank you very much, but no. The Daily Mail says it all.
 
Pepper said:
Instinctively, when I lived in your part of the world, I avoided Daily Mail & Sun. Any paper displaying tits & ass on front page just can't take seriously.
Click to expand...
I can't remember the daily mail actually ever showing T&A's on their front page... The Sun yes.. that's a comic... the Star definitely but not the DM
 
Magna-Carta said:
The fact that the robot is tethered and partially suspended at the shoulders suggests it may still be undergoing testing -- perhaps even still under development. At the very least, the software could still be a work in progress. What looks like aggressive behaviour might instead be the result of a sensor malfunction or software that isn’t yet capable of responding appropriately to having its corrective movements restrained in that particular way.

From the video, it looks to me like the robot begins to move forward slightly, then unexpectedly loses a little balance. It seems to try and correct itself, but that correction makes things worse. From there, it becomes increasingly erratic, as though it's overcompensating for its inability to properly regain balance. Perhaps made worse by the way it’s suspended -- in that the outcome of its efforts to regain balance might not be as it expected. In other demos, the company behind this robot has openly acknowledged software glitches, so this wouldn't be the first example of the robot's problems.

Does anyone else want to engage with a critical eye -- alternatively, we could just go along with the Daily Mail's sensationalism.
Click to expand...
Great analysis!
 
You must log in or register to reply here.

Back
Top