After a big night out, the drunken selfies you posted on Facebook can make slinking into work the next morning feel like a “walk of shame.” That’s why Facebook is working on an app to help you retain your dignity before your colleagues, family, and friends.
Facebook’s Artificial Intelligence Research Lab is developing a “digital assistant” that will identify whether you look drunk or sober in a photograph and ask if you actually want to upload that photo. Yann LeCun, who overseas the lab, told Wired magazine that the program would be like someone asking you, “Uh, this is being posted publicly. Are you sure you want your boss and your mother to see this?”
This enhanced facial recognition program—an extension of the technology that recognizes your face from photos and suggests tagging—will be developed using a form of artificial intelligence (AI) called “deep learning.” It’s a promising avenue in the field that ”could understand language and then make inferences and decisions on its own,” according to MIT Technology Review. Deep learning has been studied since the 1980s but has recently seen tremendous development—Android phones use it to recognize voice commands and Microsoft uses it to translate Skype calls in real time, for example.
Like any new technology, however, this development doesn’t come without controversy. Some critics said they don’t like the idea of a computer mediating internet interactions. And others don’t want Facebook to identify faces, which they can store in a database and could potentially use for target ads.
And some aired more basic concerns about allowing technology to determine a user’s sobriety based on how his face looks: