Humans at MIT Are Using GIFs to Teach Computers About Feelings

Because humans are so great at communicating emotions.

Wired recently profiled a couple of guys at the MIT Media Lab, both of whom are working together to utilize GIFs in a whole new way. While GIFs are mostly useful for trolling in the context of llama chases and the Kardashians, grad student Kevin Hu and PhD candidate Travis Rich are using GIFs to teach computers about emotions.

Although, the irony is almost stifling considering humans are, as products of the digital age, cold and soulless beings.


No, you’re right, not all of us are dementors. Jokes aside, the concept is pretty incredible even if you really are dead inside. The idea is for the human to communicate an expression into a webcam—happiness, shock, etc.—and the computer feeds back a GIF that mirrors your emotions. As excerpted from Wired:

Hu contorts his face into a rictus-like grin (“I can smile,” he mutters) and an exuberant basketball player appears on the mirror before being replaced by Snow White, who claps her hands in delight. She’s not emulating Hu’s face exactly, but when it comes to finding a GIF for every mood, she’s a fairly decent simulacrum.

This whole idea is an extension to a project Hu and Rich built together last year. They’re both the creators of a website called GIFGIF, which shows the user a pair of GIFs and allows them to choose which one better expresses a given feeling. So far, the site has drive more than 2 million votes. As Wired says, this website must be a goldmine for those studying technology paired with emotional needs.

It’d be curious to see what kind of expression the camera might look for to pull up Chrissy Teigen’s GIF above. Maybe just a blank canvas? Voldemort? Who knows.