Of lately, a lot of people have been sharing their now-and-then images from a decade ago on Facebook and on Instagram. Common users and celebrities, everyone is privy of this trend. The result is a cascade of images that display how these people looked a decade ago. While these images might have annoyed some people, it has gotten others, who believe that these images could be used by social media giants like Facebook for refining their face recognition system, worried.
Sure, the ’10 year challenge meme’, that first went viral on Facebook, looks innocent. However, experts believe that these images are giving data companies ample refined data to train their artificial intelligence (AI) systems to profile people based on their age progression.
Several Twitter users have taken to the platform to express their disbelief over the realisation. “Has anyone considered that Facebook’s “How Hard Did Aging Hit You” Challenge is just a way to refine facial recognition technology? What better way to get people to offer up a ton of comparative visual data,” Editorial Director of the Johns Hopkins University Press, Greg Britton shared in a tweet.
His views were shared author of Tech Humanist, Kate O’Neill, who, in Twitter elaborated on why the idea makes sense despite Facebook having tons of data on all its users.
“Me 10 years ago: probably would have played along with the profile picture aging meme going around on Facebook and Instagram…Me now: ponders how all this data could be mined to train facial recognition algorithms on age progression and age recognition,” she wrote in her opening tweet.
Her views, just like Britton’s, were met with Skepticism with the tweeps, some of who contended that Facebook already has all the data that it needs, so collating your existing photos won’t make much of a difference.
However, O’Neill in the tweets following her original tweets went on to explain, how images from the 10-year challenge could be used by Facebook and other social media companies for training their AI. “Let’s just imagine that you wanted to, say, train a facial recognition algorithm on age-related characteristics. You’d ideally want a broad and rigorous data set with lots of people’s pictures. It’d help if you knew they were taken a fixed number of years apart – say 10 years,” she wrote.
“Sure, you could mine Facebook for profile pictures and look at posting dates or EXIF data. But that’s a lot of noise; it’d help if you had a clean then-and-now. What’s more, the photo posting date and even EXIF data wouldn’t always be reliable for when the pic was actually taken,” she added.