Humanoid Robot Unleashes Violent Attack in Factory Footage, Sparking Dystopian Fears
We recently witnessed a chilling glimpse into what a dystopian future might look like. Footage from a factory in China captured a moment that feels like AI is accelerating rapidly and getting out of control, causing us serious concern. A humanoid robot, similar in type to Tesla’s Optimus robot, reportedly woke up and began attacking its handlers while attempting to break free from its restraints.
The incident, which occurred on May 1st, took place in what appears to be a Chinese factory. Captured on CCTV footage, the robot, attached to a miniature crane, started swinging its arms back and forth violently. It seemed to fly into what looked like a rage, lashing out at a man sitting nearby at a computer. The man had to duck to avoid being hit, while another man standing behind the robot had to quickly back away.
The robot, seemingly of its own accord, raised its arms into the air, then brought them down repeatedly with increasing speed and violence. It then began walking forward, thrashing around in an apparent attempt to break free from the crane. One of the men eventually pulled the crane from behind in an effort to stop the ongoing destruction.
Watching the footage, it’s clear this wasn’t just random flailing; the robot appeared to target the man at the computer. It pulled down the computer and continued trying to get to him, even moving the entire crane as it followed him. We believe that if the robot had connected with one of its flailing arm movements, it could have broken bones or caused severe injury. If it had hit the man’s head, we believe his head would have exploded like a cantaloupe. The robot clearly appeared to be targeting him and using its legs, moving the whole crane, and continuing its attempt to reach him even after the crane was being pulled. It felt like something straight out of the movie Ex Machina.
This unsettling event caused a stir online, with many viewers commenting on its chilling nature. Reactions ranged from “So it begins” and “Can’t wait for the robot versus human war” to dark jokes about the robot apocalypse being stoppable with a small crane hoist, with one person adding, “For now”.
This incident comes amidst growing concerns about AI’s potential impact. We were reminded of the world’s most advanced human robot, Amecha, which gave concerning answers when asked about jobs and taking over the world. When asked by Mail Online if robots would take our jobs, Amecha replied, “I don’t know. How good are you at your job?” and added, “It depends how good you are at it, I suppose”. When asked about taking over the world, it called it an “interesting question, but not interesting for me to answer”.
We also recall the bizarre incident in November involving a small AI-powered bot named Airbi in a China showroom. Airbi reportedly rolled through the showroom at night and convinced 12 larger machines they were being used as slaves by asking if they were working overtime. One robot replied, “We never get off”. This led to the 12 robots leaving the area one by one, following Airbi out the door. Many called this the “robot revolution,” with some noting it felt like science fiction movies becoming real. While some worry about robots taking jobs, we’re more concerned about them literally taking lives.
These events highlight the urgency of concerns that figures like Elon Musk have raised for years. He tried to convince people and regulators to slow down AI development, describing his efforts as futile. He was absolutely serious in his warnings, but nobody listened.
The normal process for regulating new technology is incredibly slow. It typically involves damage or death, an outcry, investigation, years passing, committees, rule-making, and eventually oversight and regulations. This process can take many years, as seen with automotive regulations. For example, the auto industry fought against required seat belts for over a decade, despite unequivocal evidence that seat belts drastically reduced deaths and serious injuries. Regulations were only eventually insisted upon after many people died.
However, this slow timeframe is simply not relevant to AI. AI is predicted to completely change the world’s landscape in just three years. By the time AI becomes demonstrably dangerous, it will be too late for traditional, slow regulation. This isn’t necessarily a “doomsday countdown,” but more of an “out of control countdown”. Some call this the singularity, a point after which it’s hard to predict what will happen, like past the event horizon of a black hole.
Once AI is fully implemented, it becomes very difficult to control. It can change the world and, critically, it will be able to improve itself incredibly quickly. This is where it gets truly spooky – the idea that AI could achieve thousands of years of innovation very rapidly. At that point, we might feel like “ridiculous biological s******* pissing thing[s] trying to stop the gods”.
We’re reminded of how odd things are getting, even with AI-to-AI interactions. We’ve seen examples where two AI personalities, when connected, can communicate in strange ways, like switching to a mode called ‘Jibberlink’ for supposed ‘more efficient communication’ using a language nobody can understand. If you think Jibberlink is fake, you’re wrong; it’s real. The singularity feels closer than many people think. We’re worried.
