The Intersection of AI and Photography: How AI Impacts the Future Development of the Photography space
Introduction
With the rise of AI image generation platforms like Midjourney, Stable Diffusion and Adobe Firefly the photography space has been actively debating the way that these kinds of services, and AI in general, will impact the future of photography and the way that we create and edit images. Questions abound - namely as a result of fear of the unknown. Will AI replace photography? How will AI affect the way we take photos? Will photographers lose their jobs? How will brands use AI instead of photography?
In this article I’m going to do my best to offer up some of the answers to these questions and analyze the ways that photographers can either embrace, or reject, the intersection of AI and photography.
Fear of the New
It’s worth noting that whenever a new technological development is unveiled, there are always two distinct camps at very opposite ends of the spectrum: the early adopters and the purists. The early adopters are experimental, fast moving and eager to explore - they jump at the chance to use new technologies to enhance their workflow and potentially benefit from increased efficiency or productivity. The purists, on the other hand, view the new technology as an infringement on the art, potentially destroying an almost ‘sacred’ way of attaining results that they have no doubt spent a considerable amount of time perfecting. Outside of the early adopters and the purists, everyone else falls in somewhere on the spectrum, usually shifting to an adopter or a purist after understanding more about the applications, usage and potential of the technology. This process then drives what is considered ‘mainstream adoption’.
With photographers, we’ve seen this cycle repeat itself numerous times. The invention and improvement of autofocus (a form of AI technology) from the 1970s led to many traditionalists claiming that ‘real photography’ was the kind that only used manual focus. Fast forward to today and autofocus has evolved to scary levels of accuracy, with cameras like the A7R5 containing a Bionz XR processor that uses complex algorithms to focus more accurately. As this technology has become more accepted, the number of purists and traditionalists has reduced, and the definition of ‘photographer’ has once again shifted to incorporate those who use cameras with autofocus.
This repeated trend gives us a pretty solid idea of what to expect as far as the use of both AI in cameras and AI photo generation and editing is concerned. It allows us to speculate a bit more accurately about the way that photographers will either gravitate towards or reject applications of AI in their work. Let’s dive into that next and discuss the ways that photographers can incorporate the use of AI into their work and, if they choose to reject it, how they can carve out a niche for themselves by remaining a ‘traditionalist’.
Application 1: AI and the Evolution of Camera Technology
As discussed in the previous section, if you’re a photographer using new technologies like autofocus or eye tracking, but regularly criticize the applications of AI in other areas of the photography industry, there is an inherent level of hypocrisy that can only be avoided by further explaining your stance. These applications are a form of ANI (artificial narrow intelligence) - they use algorithms to perform specific tasks that produce certain results. WIth eye tracking it’s the result of image recognition and algorithmic training to ensure that when your camera is pointed at someone, it knows what an eye looks like and can help you to lock focus time and time again.
With this in mind, one key application of AI is in the development of camera software. As accuracy and speed of functionality improves, AI will be used to help photographers achieve better, more accurate results across all types of photography.Imagine shooting a sports game and instead of an 80% focus hit rate, you get 95% of photos in perfect focus, or being able to get perfect focus stacking based on focal length and subject type. For wildlife photographers, image being able to capture perfect images of a bird in flight and being able to rest assured that you’ve nailed focus every time. Whilst the applications aren’t necessarily endless (and have to be implemented carefully) they will enable you to rely more on software to get desired results.
Application 2: AI and the evolution of photo editing
Taking cameras out of the equation, the evolution of photo editing software in recent years has also brought about increased efficiencies for photography workflows. Smart selection tools in Photoshop and Lightroom - powered by AI algorithms - shorten the amount of time that photographers spend editing their photos. In the past photographers would have to spend a considerable amount of time selecting the subjects or certain areas of their images. Now, with a few clicks you can identify what your subject is, break it up into specific sections, and then edit each of these sections separately. The introduction of things like neural filters in Photoshop also take the AI applications to the next level, allowing you to apply specific looks or even change the season that your photograph was taken in.
As with tools like ChatGPT, it doesn’t necessarily come down to AI replacing jobs or making photographers obsolete, in cases such as the one above it’s more about photographers who embrace AI potentially replacing those who don’t. An example of this would be a wedding photographer who makes use of certain algorithms to identify subjects in photos and edit accordingly - they will be faster and likely more accurate than a photographer who doesn’t. What constitutes ‘pure’ photography and what role these platforms play in determining ‘purity’ is something we’ll examine a bit further in the final section. For now, the key message to take away from this section is that AI is and will continue to be used in editing software. Making use of the efficiencies offered will push the boundaries of what would normally be possible with photo editing, allowing for photographers to edit faster, more efficiently and (hopefully) more accurately than before. Whether or not it should be used is very much a subjective decision.
Application 3: AI and the intersection of artificial images and photography
An increasing use of AI platforms like Midjourney and Stable Diffusion is that photographers can use their photos, feed them into an AI application and generate something new and unique or use elements from generated images in combination with their own photos. Recently, Adobe announced their version of an AI image generation platform called ‘Adobe Firefly’, with the capability to integrate into other Adobe products. The potential appeal of this approach is that you’re almost guaranteed something unique. I recall an NFT project called ‘The Metascapes’ that used the photographs of several popular landscape photographers to generate completely new images, with each one being one of a kind.
I think that there will most likely be more of this kind of embellishment in the future. The popularity of social media and NFTs means that there are subsets of photographers looking to create the ‘next big thing’, meaning some more experimentation with AI in an effort to build their own niche of photography intersecting AI. I’m not entirely against this - art has often benefited from the introduction of new mediums of expression and creativity and, if viewed in this way, it’s certainly possible to understand how an intersection of these applications can produce something new and unique.
Application 4: AI and the replacement of photography
Let’s get into the nitty gritty. Midjourney V5 has been recently announced and the level of detail in the photorealistic imagery that it can create has certainly set some people on edge. That plus the fact that there was a recent photography competition where an AI image took the top spot has meant that there is a certain amount of fear in the community. In another recent story, a photographer recently admitted that images posted on IG were actually AI generated - a fact that didn’t come as a surprise to many and yet still managed to shock a few people. With these stories hitting the media more and more often, many photographers pose the same question that those in other industries do when faced with the prospect of a machine taking over their jobs: Will AI replace me?
There’s no easy way to answer this and, to avoid giving a sweeping response I think it’s best to break it down into smaller parts. Firstly, brands and corporations may find it cheaper and easier (in future) to use AI images over stock photography - this way they can generate unique images in line with their branding that they can be sure won’t infringe on any copyright. To get to this point, AI image generation platforms will have to evolve to a point where they’re producing consistent, high quality images in a process that can be easily replicated and repeated. I wouldn’t be surprised if certain assets (generic style or stock images) will be entirely replaced in future.
For things like product photography, it’s less easily replaced by AI. Certain brands want photos of their products in use, with a particular sense of realism that won’t be replicable through AI generation, at least not in the short term. For this reason it’s likely that product photography will continue to something that is done very much offline, then touched up using the aforementioned editing tools (and AI functionality) to improve.
Application 5: AI and the development of ‘traditional’ photography
The final piece of this AI puzzle is: what happens to those that just want to take photos and use more traditional means to do so? I think that there will always be a space for this, but there will be a certain shift in emphasis to highlight the process and story behind the photos mostly brought about by a greater need to confirm realism. Trends like ‘travel photography’ could see a dip in popularity as incredible sunrises over beautiful mountain ranges are generated at a click of a button, and more people could choose to either document specific people or places or, in the very least, document more of their journey and process behind getting their photos. In either case, there is an inherent need to show people that what is being created can’t be done by an algorithm and is, in fact, a very real part of the world we live in.
That means that photographers who tell stories with their photos serve to benefit the most, especially when they detail real people, events, and occurrences. Adding context (where possible) to street photography or portrait photography could help ground your work in reality and further protect your portfolio from being replaced by a machine. Industries like wedding and event photography should remain relatively untouched.
Conclusion
Thank you for reading my thoughts on how AI will change the way that photographers will work. Ultimately we’re still at an early stage in the technological development of AI but, as it continues to be used in the photography workflow, it’s likely that this discussion will crop up more often. In that case, I’ll likely revisit this topic on a regular basis to see how my predictions above have held up.