Facebook's '#10yearchallenge' May Have Been A Huge Mistake

Courtney Stodden / Instagram

The 10-year challenge was all fun and memes until last week after a tweet moved thousands of people to worry: are we unknowingly helping giant corporations to improve their algorithms for biometric identification and age progression?

The challenge gained widespread traction on social media this month. It calls for posting two photos of yourself side by side — one from today and one from a decade ago — to show how you've changed. People are participating mostly on Facebook and Instagram, which is owned by Facebook.

Some made jokes, paid tribute to old hairstyles or drew attention to issues like global warming. Celebrities posted glamour shots that showed negligible changes from one decade to the next. (Singer Mariah Carey won this round. "I don't get this 10 year challenge," she wrote in a tweet, along with two identical photos side by side. "Time is not something I acknowledge.")

But one post went viral without featuring any side-by-side photos at all. It was written by Kate O'Neill, author of the book Tech Humanist: How You Can Make Technology Better for Business and Better for Humans.

"Me 10 years ago: probably would have played along with the profile picture aging meme going around on Facebook and Instagram," she wrote in a tweet last week. "Me now: ponders how all this data could be mined to train facial recognition algorithms on age progression and age recognition."

Her words hit a nerve.

People responded with concerns about whether they were helping the tech giant get better at identifying people. O'Neill's post got more than 10,000 retweets and more than 20,000 likes. She expanded on her thoughts in a widely shared article in Wired.

"I wondered about why this particular thought, in this particular moment, generated so much traction," O'Neill said, adding that she was not trying to stoke any panic.

Experts said the photos uploaded for the 10-year challenge were drops in a very, very big bucket of data that Facebook has been collecting for years.

"We have an awful lot of data that we're sharing all the time, and companies are collecting it and using it in various ways," O'Neill said.

Supporters of facial recognition technologies said they can be indispensable for catching criminals or finding missing people. But critics warned they can enable mass surveillance or have unintended effects that we can't yet fully fathom.

Lauren A. Rhue, an assistant professor of information systems and analytics at the Wake Forest School of Business, said the challenge could conceivably provide a relatively clean data set for a company that wanted to work on age-progression technology.

But she added that Facebook already has billions of photographs on its platform, and people should be wary of any company being in possession of such a large trove of biometric data.

Image: The fad has been significantly buoyed by the participation of celebrities (Image: Diana Penty / Instagram)

"The risk in giving up any type of biometric data to a company is that there's not enough transparency, not only about how the data is currently being used, but also the future uses for it," she said, pointing to another form of biometric data, DNA, which is increasingly being used by law enforcement to track down suspects; something many people might not have anticipated when they volunteered saliva in exchange for help tracing their ancestral roots.

"There are things we don't think of as being threats," Rhue said. "And then five or 10 years from now, we realise that there is a threat, but the data has already been given."

Like the rest of us, Facebook looked different 10 years ago. In 2009, the "Like" button was introduced, and the site unveiled a new homepage to make it easier for people to see their friends' posts in real time. Facebook also reached 360 million active users in 2009; now it has more than 2 billion.

Facebook announced it was using facial recognition technology in 2010. When people upload photos of their friends, Facebook can use the technology to suggest the names of people in the picture. It can also alert users if they are in a photo posted by a friend.

Facebook has responded to concerns about photos and privacy in the past. The company said it does not intend to help strangers identify you, and has repeatedly pointed out that users can disable face recognition in their personal settings.

Thylane Blondeau / Instagram

As for the 10-year challenge, Facebook said it's just a fun trend.

"The 10-year challenge is a user-generated meme that started on its own, without our involvement," the company said on Twitter.

O'Neill said she was glad that the meme — and her tweet — started such a broad conversation about facial recognition and privacy.

"There's a lot of opportunities for technology to do wonderful things for humanity," she added. "But I think we need to recognise the potential downsides of it."


This article originally appeared in Digital Life, The Sydney Morning Herald's home for everything technology. Follow Digital Life on Facebook and Twitter.


Comments

    1 - 99% of people posting were posting images that came from their FB or instagram archive anyway. People weren't scanning physical photos and uploading them, they were digging up photos that they had already posted to social media. So FB would have learned nothing. It can already compare photos of you from 10 years ago that you tagged yourself in.

    2 - "Supporters of facial recognition technologies said they can be indispensable for catching criminals or finding missing people. But critics warned they can enable mass surveillance or have unintended effects that we can't yet fully fathom." The upside is positive and saving lives across the world. The downside? nothing anyone can imagine. But the hysteria about AI and a skynet takeover creates paranoid articles like this one.

      It's a higher class of data.

      It knows exact dates, people likely posing similar photos too would make it easier for AI
      vs random uploaded photo where the date taken may have been different to the date uploaded or even timestamp on the file.

      Alot easier to train AI on known data or a high class or consistency of data. They can even now test their learning models against this higher class baseline.

      2. The downsides.
      What's happening in China with their ID/Citizen Score system.
      What's happening in the Five Eyes country's where private data is being misused at a huge scale.
      Snowden leaks revealed the first use of mass surveillance in the UK was to track and subvert protest groups from disrupting corporate interests.

      If you can't see it, it's not that it's not there, it's that you haven't seriously looked.

        1 - no it's not a higher class of data. You are labelling something as 10 year gap. Typically photos uploaded to facebook are uploaded within 1 month of being taken. So a) the photo timestamps on facebook are much more accurate and 2) there is a much larger pool of data and comparisons that can be done. Eg. facebook could compare a photo from 1 year ago to 5 years ago, to 10 years ago.. then compare another photo to 5 years ago and to 10 years ago and 3 years ago... The number of combinations for someone who has only uploaded 10 photos over a 5 years span is a crazy number. Imagine how many combinations/comparisons could be done with someone who has uploaded 50, 100 or 1000s of photos. Your single 10 year comparison is pure garbage compared to what they already have.

        2 - and again, you are just pulling together scary tech but not actually linking aging recognition/prediction technology with any potential problems. Once again you sound like my mum who is scared to connect to the internet because once you're connected they have your info and suddenly we are all living under china regime... As i said you are not making any valid point against this, the upside is clear but the downside is just fear and paranoia. Once again uploading your latest location to facebook or instagram and the behavior people are doing day to day is 1000 x more dangerous than re-uploading pics for comparison.

        A complete storm in a teacup.

Join the discussion!

Trending Stories Right Now