NikonGear'23
Gear Talk => Processing & Publication => Topic started by: MFloyd on October 28, 2025, 19:31:33
-
AI is more and more present in our day-to-day image processing software. As a result Adobe incorporated the latest Google product in its Photoshop latest beta version.
I'm using so called IA tools since a couple of years in my daily work. I decided to test Nano Banana's capabilities on a simple picture, but probably insolvable obstacle by classical means.
(https://live.staticflickr.com/65535/54605116570_3b621642cf_b.jpg)
(https://flic.kr/p/2rcguK5)IMG_6365.jpg (https://flic.kr/p/2rcguK5) by
Christian Vermeulen (https://www.flickr.com/photos/110731726@N07/), sur Flickr
after treatment (for demo purposes only; no practical use):
(https://live.staticflickr.com/65535/54879762007_abef51654c_b.jpg)
(https://flic.kr/p/2rBx8ge)IMG_6365-Modifier-2.jpg (https://flic.kr/p/2rBx8ge) by
Christian Vermeulen (https://www.flickr.com/photos/110731726@N07/), sur Flickr
And this by selecting the righthand grid and giving one simple instruction (prompt): "remove grid".
-
Sure works, but not for a photojournalist ;)
-
Sure works, but not for a photojournalist ;)
Please expand. The subject was rather oriented as being technical, but we can go this way up, as well.
Here my piece:
From Shadows to Pixels: AI and the Cave of Perception
Plato’s Allegory of the Cave describes people chained in darkness, seeing only shadows cast on a wall. For them, those shifting silhouettes are reality itself—until one escapes and discovers the dazzling truth outside. When he returns to share it, the others resist. The illusion is simply more comfortable.
In many ways, artificial intelligence in photography has built a new version of that cave. The shadows are no longer cast by firelight but by neural networks. Instead of physical puppeteers, algorithms decide what we see—what’s beautiful, believable, and even “real.”
For most of photography’s history, the camera acted as a witness, tethering us to the world through light and chemistry. A photo might have been framed or edited, but it still carried the weight of something that had been. With AI, that anchor has drifted. Images can now be generated entirely from text, assembled from data, or reimagined beyond recognition. They no longer record the world—they invent it.
That doesn’t make AI an enemy of truth. Like Plato’s freed prisoner, we might discover that the shadows were only one step toward a broader vision. Every artistic tool, from pigment to pixel, has extended human imagination beyond the visible. The danger lies not in the technology itself, but in mistaking its simulations for reality.
Today’s photographers and viewers must learn to see the code behind the light—to step outside the digital cave and recognize the difference between image and illusion. In that awareness lies our creative freedom: not the rejection of AI, but the understanding of what its shadows truly mean.
-
You put forward very valid points. would be interesting to learn what follows next.
As to photography itself, any picture can lie more than a thousand words. So in that respect, nothing is new.
-
… not the rejection of AI, but the understanding of what its shadows truly mean.
Well said.
-
In terms of pure technology, I'm amazed by the progression of AI image editing, except for the reversed letters on the flags which are hopelessly jumbled. :P :P :P
-
Please expand. The subject was rather oriented as being technical, but we can go this way up, as well.
Here my piece:
From Shadows to Pixels: AI and the Cave of Perception
Plato’s Allegory of the Cave describes people chained in darkness, seeing only shadows cast on a wall. For them, those shifting silhouettes are reality itself—until one escapes and discovers the dazzling truth outside. When he returns to share it, the others resist. The illusion is simply more comfortable.
In many ways, artificial intelligence in photography has built a new version of that cave. The shadows are no longer cast by firelight but by neural networks. Instead of physical puppeteers, algorithms decide what we see—what’s beautiful, believable, and even “real.”
For most of photography’s history, the camera acted as a witness, tethering us to the world through light and chemistry. A photo might have been framed or edited, but it still carried the weight of something that had been. With AI, that anchor has drifted. Images can now be generated entirely from text, assembled from data, or reimagined beyond recognition. They no longer record the world—they invent it.
That doesn’t make AI an enemy of truth. Like Plato’s freed prisoner, we might discover that the shadows were only one step toward a broader vision. Every artistic tool, from pigment to pixel, has extended human imagination beyond the visible. The danger lies not in the technology itself, but in mistaking its simulations for reality.
Today’s photographers and viewers must learn to see the code behind the light—to step outside the digital cave and recognize the difference between image and illusion. In that awareness lies our creative freedom: not the rejection of AI, but the understanding of what its shadows truly mean.
I guess that in photojournalism you are not allowed to add or subtract any objects. I know that framing the picture will have an effect, as the viewer can’t see the surroundings.
For many other photographing disciplines AI is a vet good tool.
-
I would guess many of you have seen it already, but just in case … Thom Hogan has a very nuanced position on the use of AI, basically yes and no to use it: https://www.bythom.com/newsviews/assist-versus-generate.html
-
Beware human activity is becoming obsolete, (sweet science, big tech, big money, etc.).
-
I decided to test Nano Banana's capabilities on a simple picture, but probably insolvable obstacle by classical means.
And this by selecting the righthand grid and giving one simple instruction (prompt): "remove grid".
Fascinating - thanks for opening my eyes
-
And this by selecting the righthand grid and giving one simple instruction (prompt): "remove grid".
This AI(ed) image doesn't work for me. The AI(ed) image has many people with their hands in odd positions in the absence of the fence.
"I'm sorry Dave, I'm afraid I can't do that" --HAL 9000
-
This AI(ed) image doesn't work for me. The AI(ed) image has many people with their hands in odd positions in the absence of the fence.
"I'm sorry Dave, I'm afraid I can't do that" --HAL 9000
Off course; the only purpose was to show how a complex object as a grid could be removed; no other purpose
May be you have missed the preamble "after treatment (for demo purposes only; no practical use):"
-
I would guess many of you have seen it already, but just in case … Thom Hogan has a very nuanced position on the use of AI, basically yes and no to use it: https://www.bythom.com/newsviews/assist-versus-generate.html
Completely agree with Thom.
-
The text on the two flags behind the grid looks strange after AI-modification?
On the original picture can read the letters?
-
This AI(ed) image doesn't work for me. The AI(ed) image has many people with their hands in odd positions in the absence of the fence.
"I'm sorry Dave, I'm afraid I can't do that" --HAL 9000
Yes, it is a demo how poorly "banana" works.
-
Yes, it is a demo how poorly "banana" works.
Please, enlight us :)
-
Beware human activity is becoming obsolete, (sweet science, big tech, big money, etc.).
Totally agree!