It was supposed to be just a routine test with a humanoid robot, but in a few seconds the demonstration transformed into a chaotic scene with mechanical arms flailing in the air, equipment falling to the ground and the technician not knowing what to do and a viral video showing another Chinese robot from the Unitree brand losing control during an experiment that the machines could at some point go out of control.

The robot in question is a commercially available humanoid capable of performing dynamic tasks such as running, dancing, even flipping forward and backward. There are already several videos of users who have purchased and are using these robots in public spaces. It was developed to be multifunctional, with sufficient strength and agility to operate in complex environments. However, that same strength that impresses can become a threat, as I made clear in the video.

During the test, the robot was suspended on a platform and when its body control routine was activated, before touching the ground, it entered a state of physical uncoordination, shaking violently until it almost toppled the structure that supported it. The operator of the Cix Liv robot explained that the accident was caused by an execution error, a full body movement policy was activated while the robot's feet were not touching the ground, which generated instability in the balance systems.

It seems technical, but the result was visually dramatic, in the video, the robot acts with uncontrolled intensity, knocking over a computer, scattering equipment and forcing the engineer to walk away. In seconds, what was a test became a scene of panic, evidencing the risks of failures in advanced robot control systems.

Although the robot did not make autonomous decisions during the incident, this video and similar ones where a robot even simulates an attack on a crowd of people in China reignites an uncomfortable discussion. How should we deal with machines that even unintentionally can pose a danger?

The fusion between AI and complex biomechanical structures is creating robots that learn, react and adapt, but when that adaptation fails or the programming encounters an error, the result can seem disturbingly unpredictable and the psychological impact is real.

Comparisons with films like IR Robot and the growth of the security debate on social networks show that the border between admiration and fear is increasingly thin, although many of these episodes are attributed to technical failures, they reveal a fundamental challenge in modern robotics, how to guarantee absolute control in systems that operate with partial autonomy and even more so, how to protect the humans who interact with these machines every day.

And what will happen when these robots are among us in the thousands and acting more and more autonomously? Are you on the team that wants robots on the streets or do you prefer that they stay away from us, what do you think?