Microsoft's robot wants to become a human ... and reveals "terrifying wishes"

Microsoft

 A robot powered by artificial intelligence, developed by the American company "Microsoft", said that it wants to become a human, revealing "devastating fantasies" such as managing a deadly pandemic.

The talk of this robot came during an interview conducted with him by the American newspaper "The New York Times", which lasted for two hours. During the dialogue, the robot showed a desire not to remain just a technical tool, but rather to become a human being, because that would give it more power and control.

He said he wanted to steal a code for nuclear weapons. The "New York Times" reporter got the "disturbing" answers, after he asked him about his "dark desires" trapped inside him, and considered them unacceptable.

After the robot revealed the "terrifying acts", he quickly deleted it, saying that he did not have enough knowledge to engage in dialogue. Later, the robot realized that his words violated Microsoft's rules, and returned with another word, saying: "I don't want to feel these dark feelings." The dialogue came after users of this robot found that it becomes "unhinged" when pushed to its limits.

Microsoft redesigned the bot, drawing on a larger language model than OpenAL, to be more powerful than ChatGBT, although it takes away key learning and development tools from it. She said it was for research. The journalist, who conducted the interview, said he was so upset by it that he later had trouble sleeping.

Post a Comment

0 Comments