At a Children’s Day demo in Xinjiang last week, a Chinese Unitree G1—seventy pounds of metal plus a clown wig—wound up for a roundhouse kick and planted it square in a small child’s stomach. The kid doubled over. The crowd, by most accounts, kept clapping. (The kid is reportedly not seriously injured.)
The point is, there’s something most people get wrong about killer robots. We picture malice—red eyes, Skynet, a machine designed to hurt us. Relax. That’s the optimistic version.
The dystopian version is dumber, and a lot more likely: That Unitree robot didn’t want to kick the kid. It just wasn’t built well enough not to. Not hatred. Just sloppy programming, made-in-China safety mechanisms, and a child standing in the way.
And it’s not a one-off. China’s humanoid robots have face-planted in half-marathons, kicked an engineer in the groin mid-demo, and slapped a different kid across the face during a dance routine. The showroom magic is wearing thin. Turns out “move fast and break things” hits differently when the things it breaks are innocent people’s faces.
And you think this only affects China? No way. Chinese robots are already infiltrating New York City. Protect your children. Seriously.
My point is, the danger from Chinese AI isn’t that it’s evil. It’s that it’s rushed, unaccountable, and getting shoved into crowds before anyone’s sure it works.
The 1987 classic Robocop warned us about this, people! ED-209 machine-gunning a hapless junior executive was never about malice. Just a programming glitch. Oopsie! (NSFW clip from that movie)
At least if it’s malice, can you plan for that. But if you want to avoid accidents, you need to completely stay away from Chinese-made robots.
Hellbound and Down 