Unless you’ve been hiding under a rock with bad WiFi for the past month, you will have no doubt come across the #10yearchallenge, a viral meme where people share a photograph of themselves a decade ago alongside a new photograph. For most, it’s just a harmless bit of fun to highlight what effect the unrelenting force of time has had on their once-attractive face or perhaps to show how kindly puberty has treated them.
However, some have been speculating that the challenge could have a more sinister intention: to harvest your data and train a facial recognition AI.
Facebook has already hit back at the claim, saying in a tweet to WIRED Magazine: “The 10 year challenge is a user-generated meme that started on its own, without our involvement. It’s evidence of the fun people have on Facebook, and that’s it.”
Nevertheless, the issue points to wider implications of the data we often share without blinking an eye. The hypothetical idea started last weekend with a “semi-sarcastic tweet” by tech expert and author Kate O’Neill that reads:
“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.”
As O’Neill later made clear in an article for WIRED, the tweet was meant as a flippant joke rather than a serious accusation. However, if an all-powerful social media company were to hypothetically create an AI that can predict aging, then this would indeed be the ideal dataset.
Other people have pointed out that Facebook, Instagram, etc already have the necessary data to create such an algorithm from everybody’s profile photos over the years.
An algorithm can be trained to recognize signs of aging in humans by looking at vast datasets of “before and after” photographs. Simply put, it can recognize subtle patterns – the deepening of wrinkles, darkening of rings around the eye, and all that other lovely stuff involved in getting old – and use it to predict how a person might appear as they age.
This technology could very useful too. Just imagine, police could use the algorithm to predict what missing people might look at different intervals in their life. Such technology already exists, but the bigger the dataset, the more accurate the predictions become – and few data sets could be as broad as the ones owned by social media companies. This data could also, and perhaps more likely, end up in the hands of advertisers.