Nikon D810, 80-400mm, 1/200, f4.5, ISO 12,800 |
The debate about artificial intelligence in photo editing is far from over. At what point does a photo stop being a photo and become an AI creation? Though I don’t hope to definitively answer the question here, I can at least provide a sample from the borderland.
I took the photo above during yesterday’s Central Avenue Dia de los Muertos parade. If you click on it to get a better view of the details, you’ll see just how grainy it is. That’s an inevitable consequence of cranking the ISO way up, which in turn is an inevitable consequence of working in low light (a nearby streetlight was the best I could find) without flash (too far away from the subjects for the flash to be of much use).
So when I got it home and loaded it into Lightroom, I ran a noise reduction filter on the image. Interestingly enough, the app made a copy of the original and then officially designated the filtered version as “Enhanced-NR” indicating that it had been altered using AI.
As ever, my goal is for the image to reproduce the experience, to show what I saw. The grainy photo definitely wasn’t what I was after. The noise-reduced version is a lot closer to reality.
But here’s the issue: it has also taken on an AI-view-of-the-world visual quality. If you typed “woman celebrating Dia de los Muertos” into a generative AI system, you’d likely get something more surreal than this. Still, that same sense of light and color that algorithms bring to their own products ... well, that’s what this looks like, too.
And the last thing I’d ever want as a photographer would be an image that looked like a machine made it up.