(FinancialPress) —We all know a Norman. That guy who is always finding the bleak and dark side of all situations. Some could say that their “Norman“ is a bit of a psycho in some ways.
Except for “the“ Norman himself. That one is 100% a psycho. But it‘s not his fault – he was made that way.
Norman is an artificial intelligence developed by the MIT Media Lab. His unfortunate particular case is that he was tailored by his developers to have psychopathic tendencies. The point of the experiment is to show that the data upon which AI is built is deeply significant.
Norman‘s algorithm was trained with the descriptive text of images collected from what researchers called “the deepest, darkest corners of Reddit“. Reddit is a wildly popular digital platform with multiple uses – image sharing being one of them.
After training the AI, the research team was subject to the Rorschach psychological test. His results were benchmarked against those of an algorithm with standard training. The latter saw what could be called “vanilla“ imaging in the inkblots – wedding cakes and flowers. Norman wound up with a very different perspective. He saw images of a hit-and-run fatality and of a man being shot to death.
The MIT research team explained: “Norman only observed horrifying image captions, so it sees death in whatever image it looks at.”
Norman was developed as a case study on “the dangers of Artificial Intelligence gone wrong when biased data is used in machine learning algorithms“. He was aptly named after Alfred Hitchcock‘s main character for his most-renowned work, “Psycho“.