MIT created Norman – the world’s first artificial intelligence-psycho
In the Massachusetts Institute of Technology settled “Norman” – a new artificial intelligence, which was named after a character from the cult film “Psycho” Hitchcock 1960. From the very beginning, this AI was purposefully trained to recognize scenes of violence, death, murder and suffering of people, including variants in veiled form. Materials were drawn from numerous publications on entertainment portals, where documentary photographs are mixed with memes, drawings and cartoons.
Norman is distinguished by “deep deduction”, his algorithms force the neural network to create a description, highlight key details for each scene. Then, the text information is used in conjunction with the graphic information for the analysis of new pictures. As a consequence, if Norman sees a photo of a policeman and a child with ice cream, he can conclude that this is a disguised psycho that gives the baby a poison. Norman does not care what is depicted in the picture, he gives everything to a gloomy, frightening interpretation – that’s why he was created for that.
The AI ??psychopath is designed to counter other AIs. Unlike image recognition systems, Norman can see the hidden meaning in the picture, its emotional color. But only in a negative light, and this is done for a reason. After the failures of search bots Google and Facebook, which could not resist the sarcasm and insidiousness of Internet trolls, the problem arose of learning new neural networks. And the sinister Norman serves as an illustration.
In short, he is a clear example of the concept of “what you teach, it will come out.” If the algorithms for recognizing racism or political crimes are projected by a hidden racist or anarchist, they will be deliberately flawed. And the neural network trained on them will make gross mistakes. Norman’s answers to ordinary public sound wildly, he thinks like a real psycho, but his judgments are specific logic.