look at the person in the brown coat on the right. their glasses and eyes melted together. the text above them is garbled nonsense.
and the person sitting to their left is wearing shoes that dont fit into the background and slightly overlap with the other ones' shoes. the person on the left holding their glasses seems to still be wearing glasses, and their ear is an unusual shape.
thats about all i noticed tho. pretty scary indeed.
The glasses are absolutely smudged, but that is from the image processing on the phone. Low light most smartphones try to reduce noise by smoothing the picture, often excessively.
The text above the woman is not garbled nonsense, it says “<indecipherable> is biGGer”. The upper case G’s makes it look strange, but it is cohesive text.
look at the person in the brown coat on the right. their glasses and eyes melted together.
No, that is because frames frequently have a lighter color/clear on the inside of the frame so they aren't as distracting. Tufts of hair near ears can also make things look wonky when you can't see fhe individual hairs. Blurry hands tends to be moving. This is a low quality image and the 'blending' effects are just normal potato image quality.
AI would not be able to get buy holding glasses while face in hands right,l (the black line above his ear is hair), would have screwed up the shoe logos, and a bunch of other small details.
Signs all over the image, most obvious place to look would be the woman’s face directly under the “is”. Glasses and eyebrows and nose all swirling together
Detectors aren’t 100%, blah blah, but this isn’t even pinging on any I used. Text doesn’t look AI. The image has a weird quality to it, but to me it looks more like a filter/bad camera/bad lighting than AI.
Yeah this is a real image. It's all post-processed smoothing from subtle movements captured on what was probably a 3 photo HDR bust by a camera phone. Meanwhile, some of the issues are literally artifacts from compression and the rolling shutter.
It's even better than a real photo in this case, because you don't have to worry that any depicted person is real and doesn't want their face plastered over the internet.
I can't wait until the novelty of GenAI wears off so we can resume concentrating on the message instead of the carrier medium. Either that, or until it becomes undetectable, which will probably be in 1-2 years at the current speed.
I don't agree with your view on AI but I definitely support your first point. Taking pictures of strangers and posting them on the internet has become way too normalised.
“Does this look like (specific policy) is good?”
[AI generated frowny faces]
Just wow man. The only thing trash like this accomplishes is making the movement look like it has nothing, hence the need to try and pass off slop. Enjoy your moralizing corporation worship though I’m sure next year’s perfect AI will do really great things for the world
The RTO discussion the text is about, if you didn't read the text.
What "moralizing corpo worship" do you see in my comment?
this isn't AI
Even if AI, it would be better than posting the faces of unwilling participants
and even more even if AI, i can generate images on my own graphics card, no need for corpos, TYVM, and even if i hadn't a suitable card, there are things like AI Horde where normal people allow others to generate pictures, free of charge (yes, not even paid with any ads or tracking)
You have issues, mate, but i'm pretty sure i cannot help you, better talk to a therapist or smth idc