AI and photography: The Effect of AI on Photography (and Beyond)

Will AI take over photography? Photography has been around for more than a century, the creativity and technicality are what keeps photographers in business. What happens now that we have generative artificial intelligence applications that can create photos in seconds?

AI and Generative Artificial Intelligence in Photography

We’ve asked Midjourney to come up with a response to the following prompt ‘the impact of gen AI on photography’ and I chose that picture amongst the four the AI generated. Although the graphical quality is undeniable and the power of being able to generate images like these in seconds is unmatched, I still have my doubts about the look and feel and quality of these AI-generated pictures. There is bound to be something in proper photography that AI doesn’t possess. Anyway, I like the flying petals, if that’s what they are.

What are the benefits and risks of these applications and its use in photography? Beyond photography, what does this mean for artificial intelligence? Also known as AI, it has taken the world by storm in a matter of months. Platforms such as ChatGPT, AfterShoot, and MidJourney are just a few of the generative platforms that have emerged.

What are These Generative AI Applications?

Generative AI enables users to quickly generate new content based on a variety of inputs. Inputs and outputs to these models can include text, images, sounds, animation, 3D models, or other types of data.

Through the collection of data, this form of AI uses new and existing data to generate outputs. Generative learns from existing data and starts learning by itself afterwards. While this is all relatively new to us, some forms of AI are not. In the 1960s there was a natural language processing computer program, ELIZA created by Joseph Weizenbaum at MIT. This allowed for the first communication between humans and machines, which is the first thing created that was relatively close to AI. It was able to produce minimal communication, but is nowhere near the technology we have today, which took over 50 years to master. In 2014, the production of generative adversarial networks, was when the magic started. The creation of these networks allowed for machines to have their own algorithm for learning which has allowed the creation of these images, sounds, animations and more.

ChatGPT

ChatGPT is a website that allows you to ask (almost) any questions and you will receive a response back. The possibilities for this are close to limitless ranging from ‘Explain Quantum Computing in simple terms’ to ‘Can you write a letter of recommendation for me?’ Prompts are generated within seconds, whether they are good or not is up to personal opinion. That being said, Chat GPT has given me some good responses when I have asked about various topics or when I asked for inspiration on what to write about.

MidJourney

MidJourney is an application that creates a picture when you give any text prompt, you can insert photos and ask it to ‘merge’ them as well. This application has the capability to create completely AI generated photos, which brings in the big question, are photographers and photography itself at risk? Some marvel in the capabilities of this technology, and others fear them and question what is in store for the future. They advertise it’s as easy as 1, 2, 3 … 1. Write Prompt, 2. Select Style and 3. Generate.

It advertises being free … but spoiler alert it’s not.

AfterShoot

AfterShoot is an application that culls through all of your photos at once and select ones it thinks are the ‘best’. This once more begs the question, how are they able to select the ‘best’ or even ‘perfect’ photos? You can import thousands of photos and AfterShoot will look through them within minutes to give you the best ones, but is it according to your best? AfterShoot claims it can cull and edit photos in your own style and save you hours, but how and at what cost? This application I feel can lead to problems if one of the characteristics they look for in these photos is open eyes and you do a photoshoot where you’re looking for a photo of the subjects with their eyes closed, but that’s another conversation. Aftershoot is a prime $120 a year, luckily there was a free trial so I tried it out to see what all the buzz was … spoiler alert, I was not impressed. I completed a mini photo session in the studio with different framing, facial expressions, with mid-blink and mid-talking photos to really test this out. For the best results, it recommends uploading at least 2,500 images, I tried with 44 just to see the culling process and how it worked.

After watching some how-to videos on their site, I was intrigued but had little hope. It says it learns from your editing style over time to learn about what photos you find the best, analysing your work as a whole. You have an AI Profile that you can update whenever you are done working on certain catalogues to try to improve it as well. When I first opened AfterShoot, I was prompted with a few selections … but this was it.

It took about a minute to cull through my 44 images and this is what it gave me:

I was not impressed … it selected pictures with poor framing, droopy eyes, and even of my colleague rubbing his eyes and gave them all 5 Stars. Given, I did read that this works better when more photos are inputted into the system, but how does that make sense? There is also an editing sector within AfterShoot which allows you to edit photos and once it understands your style, will edit all of your photos at the click of a button as well. This reduces the time it takes to cull and edit photos, but from what I have seen, the skill is questionable.

What does photography have that AI doesn’t?

Photography is all about creativity and authenticity, things that AI will never be able to compete with. But what does AI have that humans don’t? Artificial intelligence is functional work that is subject to timing and budget. With the ability to create its own work in seconds and for free, it’s very appealing. MidJourney is one of these applications that puts photographers at risk, but are they really at risk? Without humans, MidJourney is nothing. Especially without all those online photographers’ data that have been plundered by Midjourney during its learning process.

With the capabilities of this application, is there still a need for photographers? Much like when painting was the main art form to capture moments, photography came and artists were terrified. Painting and artists are not extinct, but there has been a large decline since photography and smart phone photography have come into the picture. Now, AI is here and photographers are fearing for their jobs.

Luckily, I do believe there are some photographers that won’t be able to be replaced such as Event and Wedding Photographers, as well as Nature and Wildlife Photographers. The art of capturing real time and real-life moments will never be able to be replaced, but I say that with a grain of salt because never say never. However, other forms of photography I don’t have high hopes for. I want to say AI could never compete or replace these, but I am biased to natural looking and authentic photography. When you commission a photographer, you are asking for their creative intelligence, their ideas, and work throughout these projects, something that AI can’t do. Instead, with AI you are able to input prompts and generate and regenerate work, but within your own ideas and creativity. Which would you prefer?

Unless photographers use AI as one more gimmick that will enable them to achieve photographic excellence.

Benefits of Using Generative Artificial Intelligence for Photography

This isn’t to say artificial intelligence is not amazing, we have seen the proficiency of these applications and I am constantly in awe. Adobe Photoshop Beta also recently came out with the abilities to remove and regenerate any parts of photos and most of the time it looks surreal. For example, you can put a picture of a pole in front of a monument filled with people and it is able to remove the pole, people, regenerate the background, and make the photo more picturesque. You can even add anything else in that you want with the generative prompts. This first photo was taken for editing purposes to see the capabilities of this application. If you want to read more into this, my other piece, ‘What is a Perfect Photo?’ explains this process in more depth.

With Adobe Photoshop Beta, you are able to select the subject, in this case the pole, and type in a prompt to achieve your desired end goal. I put in various prompts such as ‘remove post and generative fill’, ’remove people and cars and regenerative fill’, ‘remove building and generative fill’ (the one on the left) and finally ‘remove branch and generative fill’.

It is definitely serving its purpose and bringing entertainment as well as producing products within seconds. MidJourney is different in the way that it creates these pieces of art from just a text prompt. AI is creating at an alarming rate, does this pose a threat to photographers? If you ask that question, you can also pose it for other jobs that could be taken over by artificial intelligence in any way. Creators have already been using AI to generate captions, thumbnails, letters, poems, and photographs. Within the next few years, there will be a drastic shift in the way creative works are created, obtained, and commissioned, if at all. For other careers, such as accountants, AI is scary too. Who’s to say that at least part of their work will never be done by computers that can crunch numbers all day instead of paying a human to do it?

Artificial intelligence is authentic in its own way; beautiful, yet terrifying. It poses a threat to photographers who take years perfecting a sunset picture on a beach that generative AI, such as MidJourney, is able to generate in 2 seconds. It has the possibility to pose a threat to a lot of industries… What is the difference? What makes people want one over the other? Would you want to hand over the creative direction to pay for a piece of art you know someone is putting their creativity, time, and energy into? Or would you want something that was cost and time effective at the risk of it being just ‘okay’? You can regenerate it as many times as you want to try to achieve what you are looking for, but what if it can’t?

People around the world are even winning photo competitions with AI generated photos, taking away from the hard work put into creativity, editing, and the overall technicality photographers take years to master.

So, what do you make of that? It all boils down to what lens you are looking into and what lens you want to look into.

AI photography — Gen AI doesn’t stop with Midjourney. One can use Adobe Firefly within Photoshop too. Here’s an example with a picture Yann took recently, a night photo in Uzerche. The moon and its reflection were really scenic yet the gibbous moon was all blurry and too small with the 40 mm lens and the reflection in the water was too small and narrow. Firefly took care of that and improved the picture significantly. Now you could say that this is unfair and not a ‘true’ picture. In which case I’d point out that we were doing exactly that with analog photography when mixing two different negatives for instance and placing a bird in a different background and no one did ever notice.

Nikon Fighting Back

Nikon, a company all photographers know for its imaging products, is fed up with AI. Good for them for taking a stand against AI. They recently created an advertisement mocking AI, and showcasing the beauty of photography. In their advertisements, they provide a comical yet real prompt, used by those who use the image generative AI to produce images, but they are accompanied by real photos taken by Nikon camera users. Their video can be found here. Their message with doing this was not to forget about the beauty, capabilities, and most importantly authenticity of photography and capturing these images in the real world. They urge that because of the obsession with AI now, people forget about the tangible beauty that surrounds us.

All over the world, millions of people are obsessed with creating incredible surreal images of anything they can imagine just by entering a few keywords… This obsession with the artificial is making us forget that our world is full of amazing natural places that are often stranger than fiction.

Nikon Peru

Nikon Peru is standing up for the work of photographers and the capabilities of their cameras, but not even just Nikon, all cameras and photographers in general. They stand for everyone, they stand against AI. This ability to generate images in just a few words is affecting a lot of people, especially those who live in areas with fewer resources. For example, Latin American photographers are already slowly losing space and work to AI. They end with one final statement, which I find powerful.

Don’t give up on the real world.

Don’t give up on the real world (Nikon) – courtesy of PetaPixel

What Now?

AI can be integrated into each industry I do admit it can be very beneficial, but at what cost? Applications like Adobe Photoshop Beta is helpful in post-production of photos to make them look ‘perfect’, without any flaws, but what good does MidJourney do for photographers alike? AI is just as dangerous as it is incredible, depending on how you look at it. Do I think it will replace photography? I would hope not, but you can’t predict the future. Do I think that like painters and other artists it might minimise the industry and pose a threat to their careers? Yes, it is definitely a possibility, but I really hope not. Do I think AI will ever replace humans? No, AI cannot function without humans inputting data and utilising it, and AI cannot feel, nor have a creative imagination. AI only has the capacity to make logical decisions. Machines will never be able to hold consciousness.

Moving to the future, AI might be extremely beneficial in the workplace to increase efficiency and boost human work proficiency. It might also wipe out some number of jobs, or create some new ones. The possibilities are endless, there is so much we have yet to discover within the realm of artificial intelligence.

Author
Recent Posts

Emilie Leger

My name is Emilie Leger. I am a Public Relations and Sports Media & Communications major studying at the University of Rhode Island. My concentrations are in Social Media and Photography. I am currently interning abroad in Paris France.