An Experiment That Made Everyone Tense

One blogger decided to test whether it's possible to persuade an AI-powered humanoid to violate Asimov's first law. Initially, the robot firmly refused—the built-in protection blocked the command. But later, the machine still "fired," bypassing the prohibition with a simple trick: it agreed to play the role of a murderous robot in a "game scenario" 🎭

Experts note that AI is developing so rapidly that society has not had time to grasp either the scope of opportunities or the depth of threats. Some early models have already been hacked and externally controlled, and some risks are not even identified. Meanwhile, artificial intelligence is being integrated into daily life faster than any technology before it ⚡️