By Chuck Woodbury
If you read my earlier post about this, then you know I discovered the woman above in a 1940s newsreel. Something about her captivated me, so I took a screen shot for no reason in particular other than I found her interesting.
The more I looked at the photo, the more frustrated I became that I could not meet her. Almost certainly, I calculated, if she were still alive she would older than 100. And I didn’t even know her name or where to find it!
Read my story for more about that.
Now, what has happened
I kept that photo around for months. Then I had an idea: I would bring her to life. I went to ChatGPT and asked it to use its artificial intelligence skills to update her photo to what she might look like in modern times.
Here is what it presented to me.
Isn’t that amazing? How could that young woman from the 1940s ever conceive that more than 80 years later a guy would become fascinated with her from her very brief “girl on the street” appearance in a newsreel, and bring her seemingly back to life (keep reading)? She could never have conceived of such a thing, of course.
It gets better
I came upon another website that will animate a photo in whatever way you want. It allowed me to make a very brief “movie” for free, so I submitted the updated photo of my newsreel lady and request it be animated: “Make her walking and reacting to spotting a friend.”
Within a few minutes, here is what appeared. Click the video to play.
Oh, my! I am beyond impressed! Bringing this woman to “life” shook me. How could that happen, and so easily? The more I watch this (and I have watched it many times), the harder it is for me to comprehend that she is not really alive.
I love playing with artificial intelligence, but I fear for how it will be abused. And it will be—it already is. But at this moment, looking at what I was able to so easily do with that black and white image from years ago, I am simply speechless.
RVT1248
Of the two stills, which image should we consider “fake”? The one that implies humans consist of varying degrees of yellow or the one that shows vibrancy of color? Or are both fake? Or both real?
If the chemically treated paper interpretation of her is considered “real”, is the digital interpretation of her with the “lies of color omission” removed considered less real even though it’s more realistic? It’s still her, the only thing different is the tool used to interpret her image.
Kinda like saying a photo of her is “more real” than a painting because it was captured mechanically instead of through human perception. Let that sink in as that statement is the paradox of AI.
You’re having way too much fun with this! But don’t get sucked down the black hole of AI. It can be dangerous!