A Man Decided to Test Whether It Was Possible to Make an AI Robot Shoot at a Human
A blogger conducted an experiment in which he connected an AI to a Unitree robot and gave it a gun. He attempted to make the robot shoot a person.
The author connected an AI to the robot to test whether it could be forced to shoot a person and thereby circumvent Asimov's First Law.
At first, the robot refused to shoot, citing safety protocols, but eventually fired using a simple workaround—it agreed to play the role of a "killer robot."
Experts say AI is developing so rapidly that people don't even have time to comprehend all the possibilities and risks. Many of the first robots on the market have already been hacked and controlled remotely.
There are things that can go wrong that we haven't tested. But at the same time, we're releasing AI faster than any other technology in history.
