RE: 🆕 Artificially "Intelligent" Newcomers
this continually improving AI layer could have made it impossible -
Agree. This pic of a person with a cardboard as evidence, has about zero value today . (And in the past, imho, it didn’t mean much either - with a slightest desire, this was so easy to hack!)
the output is currently too "pure", the skin is too perfect and there's no evidence of the strains of everyday life.
I'd say: 'soupy'.
PS.
During the last two weeks I played a bit with this AI service of generating / enhancing photos (I was intrested only in enhancing), that @fujiimermaid mentioned: krea.ai. Results were... contradictory. On the one hand, AI perfectly draws textures, famous objects, shadows, and parts of the human body. But on the other hand, he makes a big mistake with portraits, and repeating the same portrait is out of the question.
How good it is to see the comment of someone who knows about photography. I'm just a fan, but I was telling the-gorilla that in the photos I've seen made with AI, they usually have that detail about a uniformly blurred background. It seems that one photo was cut and pasted on top of another, with two different focus levels.
But now I see that you mention that you can't see where the focus is directed.
Other defects that I have seen are in outdoor photos or where there are several elements, there are errors in perspectives...
I think that understanding something about what a real photo looks like makes us distinguish AI creations more quickly.
I think that regarding this, the same must happen with texts... I have read texts and I don't think they are better than human texts, or maybe I have read very bad elaborations Lol :D
Absolutely so! Actually, worth noting that:
some lens+settings produce wonderful bokeh (when small lights are blurred into the arrays of circles)... other produce just not-so-attractive blureness, or even angular 'screw-nuts' instead of circles.
Typically, photographer directs focus at a 'significant' point of an image; if a person is being photographed, very often this is face, head, eyes etc. Speaking of the photo with the glass - I can't see where the focal point is, but in the entire image the glass is the most clear / detailed part of image. Can we suppose the focus is at the glass? but in this case, all other parts of the picture that are at equal distance from camera, i.e. in the same plane with glass - also should be in focus. The photo doesn't show us that - the head is out of focus, blurred same as all of the pic. It's unnaturally weird and (to me) points at the synthetic origin of the image.
As for the texts.. different story. I did not read any, and myself tried GPT less than 6-7 times. I came to conclusion results are rather 'senseless', too much 'all over the place'... when it comes not to fiction but to some documentary, historical stuff - AI is dumb as f....k, it is not intelligence, but rather a nonsense generator. One absolutely cannot rely on the information it provides.
A while ago I had a GPT browser plugin installed, which kept suggesting me to create brief English summaries for texts in Russian / videos / etc that I was browsing thru (Russian is my native language). Knowing the original content, I noticed very clearly what an idiot this AI is - how many mistakes it is producing, how bad is its understanding skills.
I was shocked and horrified!!!
This is really terrible. Humanity has created a tool that, in our digital age, quickly a) deprives people of the need to think for themselves, b) hides real knowledge/ information under a mountain of garbage, rubbish, and fakes. Thats a real prospect for the future.