Imagine no longer having to worry about what you might have posted on Facebook during a wild night out.
That's the idea behind a new effort by Facebook, where a team of artificial intelligence researchers is working to develop a digital "assistant" of sorts to remind users when they're about to post something publicly, when the pictures they're putting up could embarrass them later.
The tool, currently under development, would analyze photos by recognizing who – and what – is in the shot and alert half-sober users if the picture seems questionable.
“Imagine that you had an intelligent digital assistant which would mediate your interaction with your friends and also with content on Facebook," New York University researcher Yann LeCun told Wired.
While some have expressed concerns about privacy — particularly regarding the software's ability to recognize faces — LeCun said it'll actually help protect users.
“You will have a single point of contact to mediate your interaction but also to protect your private information,” LeCun said.
For example, Facebook is developing a way to instantly notify you if someone you don't know tries to upload a photo of you without your permission, effectively giving you more control over your online identity, LeCun said.
All this comes as Facebook is reportedly seeking to up its game when it comes to using artificial intelligence. LeCun said the bigger goal is to develop software that can analyze virtually any type of information posted to Facebook, not just photos.
“You need a machine to really understand content and understand people and be able to hold all that data,” he said. “That is an A[rtificial] I[ntelligence]-complete problem," LeCun said.
Follow Jon Street (@JonStreet) on Twitter