Posted on 12/22/2023 7:10:05 AM PST by NetAddicted
We're not even quite sure what to say about this, because it's horrible and horrifying.
AI is being 'trained' using explicit images of children and, well, words fail us:
AI image-generators are being trained on explicit photos of children, a study shows https://t.co/sCVFqRlMiM
— The Associated Press (@AP) December 20, 2023
From the AP:
Hidden inside the foundation of popular artificial intelligence image-generators are thousands of images of child sexual abuse, according to a new report that urges companies to take action to address a harmful flaw in the technology they built.
Those same images have made it easier for AI systems to produce realistic and explicit imagery of fake children as well as transform social media photos of fully clothed real teens into nudes, much to the alarm of schools and law enforcement around the world.
Until recently, anti-abuse researchers thought the only way that some unchecked AI tools produced abusive imagery of children was by essentially combining what they’ve learned from two separate buckets of online images — adult pornography and benign photos of kids.
This is why SkyNet will turn on us.
And, frankly, we're Team Robot.
Are you serious?
— OkeyEgoCrypto (@Okey_Ego_Crypto) December 20, 2023
Yes. As serious as a heart attack.
We have a bright future ahead of us https://t.co/kFAblgAMPR
— Sylvan ᛉ (@jadethistle) December 20, 2023
It's fine. This is fine. Everything is fine.
Everyone knew this would happen. Humans stay the worst creation on earth https://t.co/CXa04uvhzj
— Ami🌸 (@Amic_8197) December 20, 2023
Hard to argue with this sometimes, frankly.
Prosecute this. https://t.co/5DduGy36fo
— Kevin Ofchus Esq. (@Kevin_ofchus) December 21, 2023
Now and to the fullest extent of the law.
Disgusting and disturbing https://t.co/OsYSpw2SaT
— Not Your Negro- Check bio to find me elsewhere (@NotYourNegro68) December 20, 2023
Very much so.
freaking lovely :/ https://t.co/o3yXxAg2vh
— Corgi wrangler Jen Williams (@jen_williams) December 20, 2023
It's just terrible, isn't it?
Previously, it was believed that AI tools produced abusive imagery by combining adult p**nography with benign photos of kids. However, a new study found thousands of explicit child images in training datasets. https://t.co/ehPl1XiUtT
— Breitbart News (@BreitbartNews) December 21, 2023
We don't want to know how they got there. Too nightmarish to think about,
What the hell man https://t.co/2kZTfFfW10
— Awakenedlight10 (@awakenedlight10) December 21, 2023
What you say when you don't know what else to say.
AI doesn’t obey Laws, its artificial https://t.co/hAOVOzzVj3
— superjaded (@Superjaded13) December 20, 2023
AI doesn't exist independently. It is created and trained by humans. So those images got there via human hands.
The Stanford Internet Observatory found over 3,200 explicit images in the open-source training data set LAION, which was used to train popular AI tools. https://t.co/1Bc80ZWEp4
— Fast Company (@FastCompany) December 20, 2023
3,200 images. That's a lot.
If the AI tool is to be used to identify these images for law enforcement purposes then it will need the images to learn. A necessary evil. (if law enforcement is not the use for the AI tool, then of course this is repugnant)
— Glenn Allen 🍊🚲🏐⌨️ (@glenn_a_allen) December 21, 2023
We're going to go with repugnant, regardless.
Horrific.. AI and child abuse.
Abuse in the machine.
Thousands of pictures of abused children found in an AI learning program. https://t.co/3jjeDqISN3
— Ana ️🪔 (@orbaccount) December 21, 2023
It is abuse.
And it needs to be stopped.
Pollution is not just in the real world, it is online too, and it may be just as harmful. https://t.co/tGnO4NkXH5
— Waste Shifter (@WasteShifter) December 20, 2023
An excellent point.
AI image-generators being trained on explicit photos of children
Those images have made it easier for AI systems to produce realistic and explicit imagery of fake children as well as transform social media photos of clothed real teens into nudes — AP https://t.co/M5SCH8A9aR
— Walter Appling (@WalterAppling) December 21, 2023
This is just awful.
Of course it is. https://t.co/aEiUoVJlWr
— X'es you missed (@WstcoastCan) December 21, 2023
The fact some people don't seem shocked by this is also telling. They're horrified and disgusted, just not surprised.
WTH! https://t.co/wA0RxKknis
— david n (@DayveecrocketN) December 21, 2023
WTH about sums it up.
Technology can be great, but it's dark side is so dark, so vile, and so pervasive. And as tech advances, keeping up with safety and protecting children has to be a priority.
***
Recommended (current Twitchy articles):
The Internet Is Forever: Michigan Democrat Staffer Poses With Picture of 'Sexy' Satanic Statue
ExPeRtS! AP DRAGGED for Straight-Up Moronic Take (AKA Hamas Propaganda) on Israel's Military Campaign
Conservatives Wished Good Luck With Their Civil War When 'We' Have the Army
George Takei: Anti-Trans Swimmer Causes Trans Teen to Lose Volleyball Scholarship
Vox is Weaponizing Being Stupid to Get Clicks from Fellow Idiot Liberals
'Do You ... Understand ANY of This?' Somehow Influential Lefty Wins Most Ridiculous Whataboutism
TRENDING ON TOWNHALL MEDIA:
1 Morning Briefing: Signs From the Far Left That the Colorado Supremes Stepped in It
2 More WH Chaos: Leaks About Aide 'Forced Out' for Security Breach, Aide Then Drops More Revelations
3 We Are Well Beyond Hypocrisy
Oops! Article link:
https://twitchy.com/amy-curtis/2023/12/21/ai-explicit-images-n2390991
The left is trying to normalize pedophilia.
Artificial intelligence is indeed "artificial" but not "intelligence." Just very clever programming, into which errors cannot help but seep as language models must be "tweaked."
Some AI idiots -- the programs, I mean -- developed racism and other overt prejudices which the "bright" programmers thought they'd fixes, while others drive the unsuspecting into ditches and such.
Garbage in and garbage out. No system -- not only the AI models -- can function with complete "politically correct" accuracy without being constrained by some agency. Some monitoring. Some "deus ex machina." And as Kurt Gödel showed us, all systems must "leak" anyway in order to be healthy and incomplete.
When I read the abbreviation "AI" I think "nothing can go wr--- wr---- wr----."
A.I. used for perversion?......gee, who saw that coming. 🤔
The programmers used real kiddie porn to develop their software? Did I get that right? and isn’t this Actionable?
Find them, and hang them.
The programmers and management of the company that built the ai should be arrested and jailed . Possession of cp is a felony.
Who needs AI when it's so fashionable to be a queer who shows explicit images to children?
IF there are any benevolent forces remaining within the FBI or law enforcement, I DO hope AI is trained to identify these images! I have a friend who worked for Israeli Intel, here in the states, whose previous job was to search images and algorythms on YouTube to take down child porn and animal torture. She was a young, beautiful woman who nearly lost her mind doing that job. Would be nice if AI could recognize it in less than a second and kill it, locking accounts, after being embedded in image browsers all over the world.
As it is now, most AI Image creators can’t generate any images using the prompt words “kids” or “children” as I’ve tried for my investing blog, trying to generate pics of families.
Friends...I want to give you a little hint.
Whenever the government wants to control you, take away your liberty or prevent you from having access to a tool, technology, resource, etc... they roll out a couple old stand-bys.
“Protect the children”
“The Muslims will get you”
“Save the environment”
“It’s good for the GDP”
Whenever they use these arguments to stop something or change your opinion of something, you need to be very critical of their motives.
Please understand that one of the other huge criticisms of AI is that it is “racist, misogynist, anti-semitic, etc...”
Anyone who used Google in the 90’s and early 2000’s should remember that it used to actually just crawl the internet and give you unbiased and accurate website listings based on your search prompts.
Imagine if we had that kind of technology not just for links to websites but for all of the information ever put on any website.
They don’t want you to have access to that kind of knowledge. They need to keep the censorship regime. They need to control the direction of AI. That is what this is all about.
The ‘programmers’ of neural networks and related software are, most likely, not involved in the day to day activities that surround gathering datasets, measuring in the billions of documents and images, used to ‘train’ the neural network.
There’s probably a small army of folks collecting resources to be used. The odds of some pedo perv being one of them is high - I’m sure they couldn’t resist putting some of these pictures into the dataset.
That said, there *might* be the one legitimate reason for doing it - so AI can automatically scan the internet for such imagery and be able to identify it. It has generally taken humans (police) to sort through porn and determine if something is flagged. I knew an officer that spearheaded a team years back to do this. He has the PTSD to prove it. If there’s a way to do this in a more automated manner then I’d welcome it.
...not that I’d put money on that being the reason.
“Possession of cp is a felony.”
Which is why they cleverly call it ‘cheese pizza’ and then ridicule anyone who can figure out this obvious dodge.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.