Facebook's '10 Year Challenge' Is Just a Harmless Meme–Right?
wired.com
wired.com
This was addressed in the article:
In various versions of the meme, people were instructed to post their first profile picture alongside their current profile picture, or a picture from 10 years ago alongside their current profile picture. So, yes: These profile pictures exist, they’ve got upload time stamps, many people have a lot of them, and for the most part they’re publicly accessible.
But let's play out this idea.
Imagine that you wanted to train a facial recognition algorithm on age-related characteristics and, more specifically, on age progression (e.g., how people are likely to look as they get older). Ideally, you'd want a broad and rigorous dataset with lots of people's pictures. It would help if you knew they were taken a fixed number of years apart—say, 10 years.
...
In other words, it would help if you had a clean, simple, helpfully labeled set of then-and-now photos.
It used to be the case (still is?) that the dates on uploaded photos weren't applied to the photo album. I remember having to go through holiday snaps and change the date on each from the upload date to the actual date. The images were also resized down from what was uploaded.
So, if they have the originals with the full EXIF data, I'd like to be able to use that for my old photos!
Why would it be so ? Because it profits facebook which is the only reason facebook exists. It profits them to have original with EXIF for data mining and you gave them permission to do so while also giving them the data, and it profits them to not make them available to you to save on bandwidth and processing costs.
Additionally, images used in AI are usually scaled down a lot more, 224x224 for something like resnet50. which means that they do not need your high quality original and the smaller one they generated are fine.
while I'm sure they already have a great dataset going back far, most of that dataset will be a small fragment of the world that already had internet 10 years ago, in the mean time a lot of people in other parts of the world have come online and may have old pictures of themselves
Also, people tag themselves all the time, and if you upload a photo with people, you can tell quickly that they already identify people with high accuracy, because they suggest tags for you.
That's quite an overstatement, to put it very mildly. Facebook was allowing open non-.edu signups by 2006, and the buzz around it from it's school success was immense. By 2008 it certainly wasn't seeing a critical mass of boomers and other late(st) adopters, but it was still huge by any measure - 100 million users, and growing with unprecedented speed.
People were absolutely already uploading phone pictures to FB and other sites by then; I think there may even have been Facebook apps shipping on non-smartphones by that time, it was one of the earliest things carriers used to flog data plans. I agree that "the majority of the world" wasn't uploading phone pictures anywhere by then, but then I'd be surprised if that rather high bar has been reached today either.
Their "how old" facial analysis app has been around a few years and is remarkably accurate.
* themselves
* 10 years old
https://twitter.com/facebook/status/1085675097766031360?ref_...
and rebuttal.
I know that Facebook's actions mean that it no longer deserves the benefit of the doubt but this seems like a non story that someone really wants to be a story.
Thinking about it lest say 10,000 people respond, is that even enough data to move the needle? Which photo do you use for the old vs recent? There is alot of cleanup that manually needs to be done for this to be a decent data set. Basic common sense says this is a non story.
I did my post undergrad research in 2000 in neural nets and the data sets were our biggest limiting factor, second was computation time. 10,000 data points was a huge set back then and still wasn't enough for most tasks.
It's probably many many more times that about three times the magnitude.
It might be a bit of a noisy dataset now that it's a meme.
He spun an offhand comment into full-on opinion manipulation, because we're now reading his article on the topic that not only has a clickbait headline, it seems to imply there's something more to the story, when in fact there isn't.
So now we've got all this digital ink spilled on the (entirely hypothetical) topic, and plenty of eyeballs buying it with their attention. But all of it is vapor, even at the admission of the authors.
Hence the "arctic blast" about to "ravage the east coast". Or as someone who has lived in the northeast, just winter. Or any other superlative clickbait. Everything is a crisis, everything is a disaster.
Google Photos has been able to track my goddaughter from literally her first photo (when she looked like an alien) to now (5 years later), with about 2 photos per year.
The subtle point here is that people have became so suspicious of these platforms that everything they do is observed with a sharper eye on privacy, speaking of which... I wonder how Portal is doing?
If serious, I can confirm that I saw this on reddit a few days ago too.
This entire thing sounds more like someone made a joke about "Big Brother always watching" and people without a real understanding of what's possible freaked out when they realized it is.
[1] https://www.washingtonpost.com/national/health-science/can-y...
It's just technical enough that most people who don't have a clue think that it might be right so they spread it.
Anyone who knows about data processing, programming, or AI knows that it's a very stupid idea due to easy-to-implement fault tolerance (such as random dropout) in machine learning models.
Also, is it hard to figure out the origins of a meme? Lots of them are categorized and researched pretty well already.