Recently, I presented at an Atlanta area camera club. I was shocked to learn that the members were banned from using artificial intelligence (AI) in the club’s photo competitions.
The club’s rationale
During my presentation, I talked about the generative AI tools in Photoshop because I use them. I love Guided Upright to straighten vertical lines on buildings. Sometimes, the tool leaves transparent areas that need to be filled in. I happily use Generative Fill to fix the problem.
The club says that Generative Fill uses pixels from Adobe’s servers to patch the areas. The truth is that Generative Fill looks at the surrounding pixels in the photo, uses them as a reference and creates the fill-in pixels from that. It does not pull other photographers’ pixels into my photograph.
Purity
After the meeting ended, a couple of the members asked about my use of Adobe AI tools. I explained that those tools are ethically trained using Adobe Stock. Adobe compensates photographers whose images are used to train their AI model, Firefly.
They countered by suggesting that AI was not “real” photography — that only images that are made with the camera are photographs.
Oh, please…
These members are old enough to have taken photos with film. “Is burning and dodging in the darkroom acceptable?” I asked. “Sure,” they said.
“How about using the Develop module in Lightroom? Is that okay?” Their answer was “of course.”
“So the masking tools in Lightroom are good to use?” I asked. “Yes,” they responded.
I said, “The masking tools are AI-powered.” “Yes, but,” they replied, “AI masking does not replace pixels with new ones.”
“So replacing pixels is not allowed,” I said. “That’s right,” they said.
I asked them, “What about cloning or using the Spot Healing brush?” “Those are just fine to use,” they said.
“Spot Healing uses content-aware, which is an AI tool,” I replied. “That’s different,” was their answer.
Reality check
“Okay,” I said. “Let me ask this question, ‘As AI advances, will any of this matter in five years?'” “Probably not,” they said, shrugging.
Here’s my take. Artificial Intelligence driven tools in photography are not going away. They will evolve. Their results will become more realistic. We creators will use them because they make us more efficient.
This is how human work has progressed. Cotten gins replaced hand-cleaned cotton. Steam power replaced horsepower. Calculators replaced adding machines.
I really don’t believe many objected to digital computers replacing punched card sorters. I do remember this same argument happening in the early 1990s about using Photoshop. Thirty-four years on, Photoshop is an accepted tool in photography and graphic design.
Now that Photoshop has AI tools, does that mean it has to go the vetting of acceptability all over again? And what about the club members who are denied the use of the tools in the meantime?
Early adopter
I was an early Photoshop user and have been shooting commercial photographs digitally since the late 1990s. I’m super happy that I started as soon as I did. Now, I have decades of experience in my digital craft and passion for photography.
The club members who may not use Adobe AI tools are being kept from learning new tools. In my thinking there is no way this is good or right.
My suggestion to the members of the discussion was to add an AI category to the existing B&W and color categories.
Progress will progress no matter what.












Your comments are spot-on.
For me, the use of AI tools comes down to what is the purpose of the photograph. If it’s art, then I have no problem with it, but equally I have no problem with competitions that ban it; at that point it comes down to personal preference. Where I would have a problem with AI tools in particular, but other tools too, is where it is used to lie to the viewer. Deep fake images are perhaps the most pernicious example of that, but removing an ‘offensive’ door mirror from an advertising photo of a car, or replacing a perpetually… Read more »
Agree… In competitions you need an equal playing field. If you are now creating a great landscape, sky and foreground — and it comes from an Adobe library… Ok, then its just a plain old art show in the park or an Instagram like… The art of the photograph is lost. Neither is bad — but they dont mix.
This reminds me very much on the times when color photography was not regarded as artistic becaue the colors were produced by a random process not an artistic one (not that I lived at the time). I also disagree with Davy who said that replacing the sky for a holiday brochure is OK – here we are already on a slippery road as it may be regarded as fraud especially when the sky is perpetually smog laden. Anyway AI can help many of us creating the picture they had in mind without being a Photoshop master. Thanks for an interesting… Read more »
People should use the tools that best support the work they do and the message they want it to send. Don’t like a tool? Don’t use it. It’s that simple.
Progress marches on. So sick of dinosaurs trying to hold us back, in politics, and in photography.
the problem is in completion — you need an equal playing field.
But what if the AI completely creates a photo?
I asked it to create a shot of a man Playing blues guitar in a smokey Bar and 30 seconds later I had the shot – of a man who doesn’t exist, in a Bar that doesn’t exist…
Exactly – its bogus… Its just plain old street art — not photography…
When Adobe or other software makers use the term (currently anyway) generate or generative fill, that means they are going to their database of images and using parts of them in your image. You seem to think that small fills are not happening this way, but you cannot know for sure. Try Expanding more one side with background detail, and it easily becomes apparent that things are coming from elsewhere, and are someone else’s photo parts. (Indeed Adobe has explained this as true, even for small fills.) Now if you are okay with doing that, it is okay, but my… Read more »
Except that is not how it works. There is not a data base of images that it pulls from. It is trained on images which it equates with labels (dog, whatever) and those are converted to noise. When you ask for a dog, it makes a dog, based on what it “knows” a dog to be, not on a database of images it uses. And with guided imagery you can dial it in more, esp. if you use your own photographs, or build a data set from your own photographs.
As a photographer, I can’t get used to AI being a part of my editing, especially when submitting to a photography competition. My feeling is that there should be a definite dividing line, 2 categories, whatever, but if I take the photograph, it should be totally my work, my talent, my expertise, and on the other side of the dividing line is the AI person who sometimes is not a good photographer, but someone who is good at using something false to improve their second rate image. For me, the two don’t go together, but have to be separated.
We had PhotoShop, now it is AI. So what! The result matters. In 2 years everybody will ask, what a discussion this one was.
NO, AI IS A GREAT HELP. The photo still has to be made by the photographer. Composing, the best moment and so on, is more important than AI.
I use lumina XE and since I am not the best photographer and AI help me improve my photos that I have not mastered. Plus I find sky replacement improves the look of my overall photo.
You are so correct Kevin! What, if anything, is “real” or “pure” about the art and craft of photography? The key words here being “art ” and “craft” Thank you for an excellent reminder of these two facets of photography.