I've read a research paper where they trained an AI to generate new human faces.
These faces are fake, fabricated, and have never existed.
But they're indistinguishable from real ones to any observer.
Imagine casually surfing through these faces and finding the face of a friend?
It's possible there's been some malpractice from the researchers (obtaining photos from the internet),
or that there is a finite number of templates that faces have to fall into in development (which could explain lookalikes),
or that there is a finite number of templates that faces have to fall into in development (which could explain lookalikes), or that you are simply seeing shapes in the clouds.
OR that your friend isn't real!
Now imagine finding your own face.
Again, the malpractice, or the finite templates.
OR that you aren't real.
But this is different from the friend situation. You'd exist enough, be real enough, to see your own face, but not be actually real.
What would that even mean? What would it be like? What would have to be true in order for that to make sense?
The paper is real; those are the faces.
The questions are mine.
And as usual, I've written a short story about it :)
It's available for pre-order on Amazon. Consider reading it for more on this! (link in comments)