This is a most excellent place for technology news and articles.
Hpw do you run ads for child abuse?
Instagram has been profiting off paid adverts that promote CSAM in India
OK...
How did we get here?
Earlier this year, Meta announced it was trying to reduce its reliance on third party human moderators by — you guessed it — switching to AI.
That's what they said would happen...
How bad is it?
The videos promote child abuse and invite you to join a Telegram group to have access to it. So, that's bad, and also sounds like a honeypot.
Programmatic Advertising. Basically a automated auction for advertising spaces – the more visitors, the more expensive the place — with practically no oversight (mostly through Google and Meta) of who takes part. Which is also a reason, why this practice has a problem with malware (malvertising) and fake news and why it should be illegal, but you know how it is over there.
I want to be clear I’m not promoting Meta’s decision here by any stretch.
There is an interesting problem with human moderators though. Once the entire population of Earth is posting on social media, are there enough humans to ever moderate it? You might say oh sure, because the average person might only post occasionally and human moderators could be dedicated full time so they’d outweigh users by factors. But that kind of ignores influencers who also generate content full time.
I really don’t see how there will ever be enough humans to do this right. Or if there are, they will have to number in the hundreds of millions to keep up. The bit about all humans posting is not really hyperbolic at all. Zuckerberg already has 1 billion of us.
There is an interesting problem with human moderators though. Once the entire population of Earth is posting on social media, are there enough humans to ever moderate it?
Yes, because the community itself can be the largest part of its own moderation. As Reddit has proven, they do not all have to be paid, and people who run their own groups would be happy to take more of that moderation work on just to ensure their own viability on the site. There can also be local moderation as well as top-down moderation: it does not all have to be sweat shop workers in the third world. Meta could do far more than they are now to limit sockpuppet and astroturfing accounts, and accounts that are not actually tied to a human being, all of which would have an oversized effect on cleaning up the filth, but they don't.
Rather, time after time, year after year, people tell whoever is looking into [whatever the current scandal is now] that they reported this content and Meta came back with "Nope, we have looked into this and this content does not violate community standards." What if Meta just took action on the community reports they already receive? Meta will rapidly and without recourse take down content that involves, for example, classic art that involves nudity, but double down on ignoring actual CSAM, as in this case.
Compare and contrast that with Meta's vigorous policing of content that does NOT promote their agenda. Put those together and it paints a picture of a corporation that is very much involved with its ability to manufacture and sway public opinion, but has no interest otherwise in cutting off anything that makes it a dollar.
There are all kinds of helps and fixes to what has become an extreme moderation problem, every single one of which Meta has refused to engage despite having profits in the billions. Billions. You could hire an entire country of moderators with a fraction of that, but they don't simply because they are not interested in any of that.
No one is on a Meta site to be served as a customer; they are there to have their data collected and analyzed for better efficacy of the propaganda and advertising Meta pushes their way and then sold on to other data aggregators or even the government, and to be used in any way that Meta likes without recourse to the user. And if a kid gets trafficked, or killed, or kills themself because of Meta, well that need not force any real change either. Meta's gotten really good at making concerned noises about that kind of thing without ever changing their own part in how it happened, and continues to happen.
Or to put it another way, in today's model, Meta sees itself as entirely separate from its userbase, and unable/unwilling to distribute any of its power among its users to make it better for everyone. But look at Lemmy, where there is little separation between instance admins and the general userbase: everyone can help moderate from the ground up. There is no money in moderation here at all, yet we have not descended into 4chan. How is it that Lemmy can do adequate online moderation (and Reddit before it) for free without a single dollar, much less without the billions of dollars available to Meta? Its because on Lemmy the community itself is the primary moderation tool, and while they talk otherwise, in reality Meta refuses that community paradigm almost entirely.
There's another factor as well: it is far easier to moderate an online community that has always kept its standards high from the start than it is to come around and clean house when it all starts feeling a bit Voat and 4chan. Had Meta been adequately moderating all along, it would never have gotten this bad to start with.
TL;DR: Effective online moderation only looks like an impossible task when it suits a company to not even try.
I'm sure it's still possible for humans to moderate paid adds content, its a small part of the overall traffic volume.
Fair enough. But I think we’d be just as upset to see pedos organizing in unpaid content.
It’s absolutely possible to moderate. But doing so would require more payroll than the tech companies are willing to spend. So instead, they simply let CSAM ads run rampant.
Also, exposing the moderators to extreme content like this is traumatizing.
Yes it’s a thing with more than one challenge.
What would 200 million moderators cost? At a modest salary of $50k annually it’s in the trillions. They literally couldn’t pay that.
Facebook has the money to hire enough moderators, though. They're starting to not grow any more, but content moderation is one of the last things that should get cut instead of the first. When Yahoo started dying*, moderation was also the first thing to go.
Like a canary in a coal mine.
* it's technically not dead, but it's basically dead. People keep telling me Yahoo is inevitable, and its continued existence is evidence of this. /s
I agree with everything that you’re saying and yes they have a lot of money. Do you see what I’m saying about how this could require millions and millions of moderators at the kind of scale they operate? A significant portion of humanity is on social media.
Meta earns about $200 billion in revenue and $60 billion in profit. How many mods would $60bil actually buy?
Let’s work with a total cost of $50k per moderator. This is very conservative since only a fraction of that would be salary (the rest insurance, HR, management, taxes, equipment, etc).
If they spent their entire profit on mods, they could hire 1.2 million. Is that enough to moderate several billion? Maybe.
Now should we consider costs they could actually sustain? Taking their entire profit to zero is, understandably, not possible for any company. I think this is where most people just wave their hand and say “who cares” or “fuck em” or just assumes they have infinite money.
But to seriously consider this… it’s just a very big challenge for anyone to pull off.
Zuck is one of the richest people on the planet. He has the kind of money that could track the original source of the advert and pay to have them publically executed, then pay the fines and court fees to publically show why it was justifiable.
They don't moderate these things because they have been facilitating it.
Welcome to capitalism.
No worries people, nothing new from Scumbag Land.
Move along now, hurry up, tut tut!
AI being used to PROMOTE CSAM? This is MORE IMPORTANT then LIFE!
-The US Government which Argued CSAM Generator GROK is MORE IMPORTANT then Life!
