Can Machines Learn to Scare Us?

Why do we get scared? A lot of it may be due to things like cognitive dissonance or perhaps even our brains are unaware of exactly the situation going on. The unknown seems to play a common role. Whether it is through novels like Mary Shelley's "Frankenstein" or Bram Stoker's "Dracula," or more modern day horror movies like Kubrick's "The Shining," a cornerstone of culture seems to be fascinated with striking fear into others and ourselves. However, as technology advances and with the emergence of artificial intelligence, there seems to be one central question—can machines learn to scare us?

And that is exactly what MIT plans to do with Nightmare Machine.  Nightmare Machine takes certain algorithms and calls for public action across the internet to help this algorithm learn exactly what is and isn't scary. If you do decide to assist the AI, you will be given a spooky or less than spooky face tasked with choosing whether this face is "Scary" or "Not Scary"

Upon the completion of ten of this very simple test, you are rewarded with the "special hall of faces" which can then be easily tweeted! I think what MIT is doing is really interesting and is paving the way for what could only be deemed horror AI. Being able to be a part of this process is something that I think may be pretty cool and something I'd definitely suggest checking out.