Generative AI is transforming how humans use dating apps and even spurring real people to romance AI chatbots.

Uhane said:

The uncanny valley runs deep but it is not very wide (so far)

That isn't the trend I'm seeing, unfortunately.

This photographer had a large successful portfolio on agencies all of their own photos but now they are only/mainly using AI I believe.

If you want to take the link, you will see many realistic looking AI created images:

https://stock.adobe.com/search?creator_id=203855208

I should be giving MI any help but -

There is always a tell. It may be subtle but there are perspective inconsistencies,vdead eyes, suggested rather than overt features, faux pas in the shadows, the wrongness in the way light shines on objects, or a sort of murky global lighting to avoid having to deal with it,  impossible anatomy, impossible scaling, "silly puddy" skin tone, or too consistent skin pores or–the most glaring–inconsistencies in the focal lengths and depths - or the lack thereof.

I cut my adult digital life in Photo shop from when it first came out in beta, so it is easier to see, perhaps, after all these years. The machine intelligence's efforts use all those same tropes.

For web use these pictures are also boiled down to 72PPI and this becomes even more obvious. It's down in the pixels and how they differ from source to source as well which wont be there unless they make something up.

(I wonder if I should be saying all this in a public forum where MI can take notice.)

•••

Maybe I'm just be good at seeing when things are in the uncanny valley, or generally wrong or different.  When I go to the check out - I cannot remember (or don't think to register) the color of the cashiers eyes (I still struggle with this)  but I can tell if they've plucked their eyebrows in a new shape or have gotten new nails, haircut, clothing.

The first time I saw someone who'd had cataract surgery I started with panic and could not look even in their direction - really thought they were not human: In the pupil there were dancing hard lights catching reflections.

Lets look at this picture

Note the skin is wrong - putty face smooth- and all the same tone, also a bit out of focus. And  the teeth - are too perfect and also all the same color, and the hair: that is just the "oil painting" filter- and rather poorly executed at that - so in looks like that stuffed animal fur only greasy. I also suspect they rip from older, public domain photos that have actual film grain that the MI eliminates with smart blur filters etc. I'll shut up now.. sorry to go on and - my, look at the time!

Ive over stayed my allotted 20 minutes! Sorry Debbie.

Thank you Uhane .

That was very interesting.

thanks back, Dear Debbie, I wonder sometimes if I am being too .. pedantic or just helpful. But I am among peers so, thanks you.

Can you see now? the difference a bit better? AI cannot grasp gritty detail and homogenizes everything to avoid it.  

I thought, as you are a photographer yourself you would understand. The 2 photos I posted on creative space are a sort of answer to the AI milquetoast dead-ness of an AI image's data driven concerns. 

thanks back, Dear Debbie, I wonder sometimes if I am being too .. pedantic or just helpful. But I am among peers so, thanks you.

Can you see now? the difference a bit better? AI cannot grasp gritty detail and homogenizes everything to avoid it.  

I thought, as you are a photographer yourself you would understand. The 2 photos I posted on creative space are a sort of answer to the AI milquetoast dead-ness of an AI image's data driven concerns.