Even better than the real thing : In the world of image AI
(with apologies to U2)
“That’s fake fog.”
“That photograph has a fake sky, blended from somewhere else. “
“The skin tones on that portrait are completely plastic. Are we looking at a human face or Barbie?”
“The color gradients on this image look so unnatural. “
As we flip through photographs on our Facebook or Instagram feeds, 500px or the Google Photos feed on our Chromecast, we cannot contain our frustrations around how fake the world of photography has become.
How naively and transparently fake.
Sky replacements. Addition of fog and clouds. Conversion of morning photographs to evening light. Adding the ray(s) of God from heaven. Object cloning (not just to cleanup, but to add to the story). Removing every iota of a skin texture or blemishes from a face, thinning them down to supermodel proportions, and brightening smiles to make your dentist feel irrelevant.
By a rough unscientific count we think more than 70% of our feeds have had serious “fakery” done to them — we are not talking about the dodging, burning, saturation, exposure corrections kind — we are talking about “why even bother calling this a photograph” kind fakery.
With machine learning and excellent edge / object detection algorithms, this is becoming even more accessible — check out what Google Photos or Landscape Pro can do to your photographs. Maybe this is the new norm. Maybe we are done with the age of photo journalism and now its all part of the large spectrum of art — and whether it is kitsch or fine art is up to the viewer’s discretion, and the number of likes.
Originally published at rowdy planet.