I watched a Casey Neistat video yesterday about AI, and it made me wonder how this technology will be helpful and harmful for photography. AI is not going away. Photography isn’t going away, either—at least not anytime soon. What kind of impact will AI have on it? How do we move forward?
Let’s start with the harmful. Below are some quick thoughts on why AI is bad for photography:
Devaluation of human craft. AI images lack human story, skill, and effort, yet still compete with actual photos, diminishing appreciation for what is real. Art requires human effort; fake art is faux-art.
Job displacement. Photographers, photo editors, and even models will—and have already begun to—lose work as AI tools and generative systems become cheaper alternatives to paying people to create real pictures.
Authenticity issues. AI blurs the line between real and artificial, contributing to a distrust in images. Oftentimes, our first thought is: is this AI?
Loss of style. As computers take over what once was individual choices, everyone’s photos will start to look alike. There will be fewer and fewer new aesthetics.
Ethical & copyright concerns. AI has been trained on real works by real people without their consent or compensation. Cases currently being hashed out in courts are likely to determine that companies illegally trained their AI models, but the damage has already been done, and the little guys—that’s you and me—will not get even a penny for it, while Disney and other large companies will get large settlements.
Fundamentals skipped. When cameras and software can do everything for you, technical skills and creativity will become less important and less prevalent.
Cultural & artistic dilution. Photography’s value as a historical and artistic medium will erode when infinite machine-created images overshadow real human moments. AI will increase the total number of consumable images available by leaps and bounds, and non-AI images will be more easily lost in the massive pile of AI slop.
Energy consumption. AI takes a lot of power and resources. By 2030, it’s expected that up to 9% of all electricity generated in the U.S. will go to power AI. That’s shocking!
Now, let’s look at the positive. Below are some quick thoughts on why AI is good for photography:
Powerful tools. AI can be used to remove distractions by handling tedious tasks, freeing photographers to focus on other things. A very smart friend of mine told me once, “I’m not afraid of losing my job to AI; I’m afraid of losing my job to someone who’s better at AI than I am.”
More proficiency. AI can help you do more. What once took a team of people can be done with just one or two.
Accessibility. Photography has never been more approachable for beginners, those less experienced, those without expensive gear, and those lacking technical knowledge. And it’s only going to get easier and easier. Literally, anyone and everyone can be (and is already) a photographer.
Authenticity revolution. As AI images and AI altered photographs become more and more common, the natural response will be a yearning for what’s real. Authenticity will become even more valuable. Unedited digital photos and film photography have been gaining popularity, a trend that will continue to grow and grow as a pushback against AI.
My view is that AI is largely bad, but it can be good. It certainly can’t be ignored. So how do we navigate such treacherous waters? I don’t use AI often, but I use it sometimes. Like with most things, moderation is key. Use AI wisely, and not for everything. Let me share some recent examples of how I used AI.
The most obvious is the image at the top, which isn’t particularly good. It illustrates the topic of this article well enough, though. I was able to generate it in a few seconds with just a few words typed out. A few months back I used AI to animate a small section of my YouTube video to help tell a story. Those are pedestrian uses of the technology, and not anything to get excited over.
Where it really came in handy for me is back at the start of summer. I found a deal on an RV rental. This particular unit needed to be relocated from Miami to Boston, and it needed to get there in two weeks. Anyone willing to do it would get 75% off the rental cost. Driving up the East Coast in an RV is something that my wife and I had talked about doing for a few years, but it just wasn’t practical. Suddenly, this discount was available and we happened to have a hole in our calendar, so it was like fate. Even the airline tickets were surprisingly affordable to Miami and from Boston, especially considering that it was a last-minute purchase. The only problem is that we barely had time to pack, let alone plan an entire trip up the Atlantic coast. So I used AI to plan it for me.
I told it when and where I was starting and ending, and let it choose the whole route, including where to park the RV at night. I made some RV park reservations (AI even provided the websites and phone numbers for that), and soon we were on our flight to Miami. While not perfect, AI did a better job at planning the trip than I could have, and it did it in seconds, where it would have taken me hours and hours. I was surprised and impressed. I would definitely do that again.
In fact, I did do it again, but on a much smaller scale. I visited the Great Smoky Mountains for the first time a few weeks ago, specifically for fall color photography. I wasn’t sure where I should take my camera and what time of day might be best for those locations, so I had AI do that research for me. It was tremendously helpful, and saved me probably an hour or two of research.
When I’m creating new Film Simulation Recipes and I’m researching films stocks, I have used AI to help with the investigation. While AI is great for research, it isn’t always right. Sometimes it gives a wrong answer. Sometimes, if you ask it the same question twice, you’ll get two very different answers. These instances of inaccuracy and inconsistency does call into question if AI is trustworthy at all. It seems to be moderately trustworthy, and requires double-checking its responses for anything remotely important.
Different AI models are better at various tasks than others, but none of them are good at creating Film Simulation Recipes—at least not yet. Initially, AI was quite terrible at this, often giving nonexistent or nonsensical settings. It’s better now, but still quite mediocre. It doesn’t have a good grasp on what the practical consequences are of camera setting adjustments. If you ask it to create the exact same look more than once, you’re sure to get two notably divergent answers. As someone once told me: AI will often give a better answer than a complete novice, but rarely a better answer than an expert. You can see the progress over the last couple of years, and I’m sure there will be continued improvements, but we’re not there yet when it comes to Fujifilm Recipes.
I think that navigating AI requires using it. If you are not at least somewhat comfortable with the technology and proficient at it, you might (as my friend worries) lose your job to someone who is better at it than you are. But moderation is still the key. Use it when it might actually be helpful. Don’t use it for everything. In fact, using it for everything might be to your detriment.
Some recent studies suggest that over-dependence on AI can reduce critical thinking skills, memory, and creativity. It might also make you lazier and feel less motivated. A quick Google search reveals headlines like AI Is Making You Dumber, Yet Another Study Finds that AI is Making Us Dumb, Science Shows AI is Probably Making You Dumber, Studies Prove It: AI Makes Us Dumb, and so many more. The key seems to be how you use it, and clearly less is more.
AI can be a powerful tool, and using it affectively can be a great benefit. It can save you a lot of time and increase productivity. It can help open some doors that might otherwise be impractical to open. But there’s also a negative side to it, and overusing it can have negative side effects on your mental health. It seems tricky to navigate. I think, though, that the technology will become more integrated and more seamless, and soon we’ll be using AI without even realizing it. Perhaps, in a few years, things will get sorted out, and it won’t be such a treacherous path. In the meantime, consider moderation as the safest route through this tricky time—not ignoring AI entirely, and not using it for everything, either.
The cost in water should be mentioned too. This is horrific. Available water and energy is already lowered around big data centers for AI. https://www.bloomberg.com/graphics/2025-ai-impacts-data-centers-water-data/
That’s a good point, too. They’re building some in the dry AZ desert, which just makes zero sense.
Hi..
I get annoyed when I see a Ai photographs that people think are real. Stunning, fantastic beautiful are the words they use. To me it’s a fake image generated by a program. It can be seen as a form of art but not photography.
I had a landscape (framed) photo that use to sell well, recently I got a comment why should I have that on my wall when I can have a similar image with a dragon flying by or an erupting volcano in the scene. My reply was because it’s not real.
The Ai craze can work in reverse, I placed an image taken on my XE-4 in an online photo contest. It got denied as we don’t except Ai? I had to prove it was not Ai. They said sorry and excepted it but left me angry.
A photographer paints with light, goes out and finds the photo and then with their camera, lens & skill creates the image. I don’t see Fujifilm selling Keyboards. LOL
I grew up with photography, as an old timer from the film days in the 1970s.
Ai has it’s place in editing but when it lies and tries to take over from reality then its going to far.
I gave up on my Facebook page due to it smothered in Ai rubbish and fake Ai bot accounts.
We need more laws and rules for this Ai craze before it’s too late.
I think anything AI created or manipulated should be clearly marked as such. Not that it’s difficult to tell, but just for the sake of transparency. I’m sure some people will still prefer AI slop even though it is fake, but at least we’d be on the same page of what’s real and what’s not.
I basically agree with your opinions. It’s worth considering that, for environmental reasons, there are already plans to put all the hardware into orbit in space (free energy and cooling), and I believe that many aspects of current geopolitics are based precisely on the space race for these reasons as well. I agree with you that it’s a technology that can’t be ignored; in my field of work (healthcare), several activities that used to require 3-4 hours of a doctor’s time have already been reduced to little more than a minute. And this is precisely one of the aspects that are worrying: given that it’s a technology that impacts the middle-upper classes (hairdressers are safe :-))), the ones that contribute the most in terms of taxes, and that it’s essentially a technology that doesn’t replace “old” professions with “new” ones (as have been the case with past innovative technologies), there’s a serious problem with the sustainability of social systems. I imagine this process has already begun in the photography field. Perhaps only authorial photography, the kind that “narrates” in a new way, will survive, because AI only works by looking “into the past”; however “generative” it is, it’s not “creative.” Obviously, these are all my personal opinions, food for thought.
Thanks for sharing! It’s definitely something to chew on.
There was an interesting review from Emily at Micro Four Nerds, about an AI-enhanced camera called Caira, which used a m43 sensor. You could do things like change the Ricoh GR in the picture to taxi-cab yellow. Or change trees to fall foliage, or put the cherry blossoms in bloom. Fake nonsense, and Emily was pointing out how it’s all kind of a nightmare.
And then I kept thinking about some of the fantastic computational photography tools in Olympus/OM-D cameras. Live ND stands out, but other sorts of in-camera focus stacking and compositing. There’s a lot of powerful stuff, and useful stuff. Maybe there’s more space for expanding those ideas. Maybe defocus stacking, instead of focus stacking–the portrait mode from a phone, but done with much more serious glass. And that’s not what AI photography is doing.
So not only is it an ethical nightmare, built on stolen content, with the intent of displacing those it robbed. Not only is it an ecological nightmare with electricity and water usage. Add in the core problems of bias confirmation and over-relying on an optimization solution. But it’s all just so dumb. Make fake stuff up, rather than actually doing something useful. Some idiot in charge decided that it made sense that the first thing to do was flashy useless stupid things and made-up nonsense.
And I do think photography is particularly incompatible with AI. The point of a photograph, for a lot of people, is that it’s the thing itself. It was a huge revolution in the press world when we could have actual photographs, instead of just an artist’s rendering. Granted, those drawings can have merit in themselves, but having THE REAL THING was a huge reason why photography first became popular. Replacing it with an ~unreal~ thing is not just against the grain, but destroys the entire point.
I fully agree with all of this. A potential good use of AI is AF and AWB improvements (I guess time will tell if it actually does). Creating a fake image of a robot with what might be a camera at a vague representation of the Grand Tetons is a poor use. But it’s certainly being used in even worse ways. If AI kills photography, it will be the fakification (hey, a new word) of photography that kills it—how precisely AI is utilized. It can be for good; so far it’s mostly not. Thanks for the thoughtful comment!