At what point does our brain perceive an image as depicting a real event or a real person? When do we, as viewers, know for certain that an image is a photograph versus one generated by AI? Below, I present a few images.

These are alternative visuals for a story I posted as part of a compilation of fictional cannibal tales:

"The Dark Secrets of Celantanis – The bond between the hungry sisters"

https://boards.guro.cx/lit/koko.php?res=30684&page=1

Obviously, they are all AI-generated. You know it, and I know it—I certainly do, since I created them. But what about people who don't usually engage with AI imagery? In my stories, I tend to dial back the level of realism.

However, here I want to demonstrate that the degree of realism in AI-generated images can vary significantly. Which of these images would the people around you classify as real?