"Upgrade" Blumhouse

Maybe something like Upgrade is a lot closer than we think because scientists from Massachusetts Institute of Technology (MIT) have created a psychopathic Artificial Intelligence (AI) using Reddit images.

They call him Norman after Norman Bates from Psycho. The experiment according to IFLS was to supply his algorithm with certain data to see how it influenced his outlook on life.

“Specifically, how training an algorithm on some of the darkest elements of the web – in this case, images of people dying grisly deaths sourced from an unnamed Reddit subgroup – affects the software,” reports IFLS.

Norman’s “brain” can comprehend what he’s seeing and write down a description. MIT created another AI, this one was fed only “more family-friendly images of birds, cats, and people.”

The two were then given an inkblot or Rorschach test. This therapeutic tool is made up of images which are interpreted by patients to gauge their emotional and mental health state.

The results were striking.

For example in the image below, the standard AI described it as “A couple of people standing next to each other.” But our friend Norman said it was, “Man jumps from floor window.”

And it only gets more disturbing from there. In the next one the standard version said it saw “A black and white photo of a baseball glove,” Norman wrote, “Man is murdered by machine gun in daylight.”

Next, “A black and white photo of a small bird,” said the standard. “Man gets pulled into dough machine.” Yeah, Norman’s seen some things.

The conclusion is that the data Norman ingested mattered more than the algorithm itself.

“Norman suffered from extended exposure to the darkest corners of Reddit, and represents a case study on the dangers of Artificial Intelligence gone wrong when biased data is used in machine learning algorithms,” the team, who are also responsible for the Nightmare Machine and Shelly, the first AI horror writer, said.

What the scientists behind Norman’s existence proved was that other algorithms can be influenced by negativity too. They point out Microsoft’s Chatbox which was deactivated because it kept speaking in trollisms: “Hitler was right” and “I f**king hate feminists and they should all die and burn in hell.”

Hope for our Reddit worshipping Norman is not lost though, you can help his algorithm recover by taking the inkblot test yourself.

Source: IFLSCIENCE

1 COMMENT

  1. Why the hell make an AI PSYCHOTIC? that’s moronic…things can an do go wrong! Not safe or foolproof!

LEAVE A REPLY

Please enter your comment!
Please enter your name here