AI in Photography: Creative Tool or Cataclysm?
Just as I arrived back from holiday the news broke that Adobe had just released their Generative Fill tool in the beta version of Photoshop. You can read more about the tool at the link, but in a nutshell it allows you to use AI image generation technology to replace or fill parts of your image.
My social media feeds, such as they are, reflect my general interest in photography and, specifically, an interest in club photography. The blogs and YouTube channels were awash with comments which, broadly speaking, broke down into two camps: “Isn’t technology great” against “‘tis the end of days”.
Given that I’m currently president of my local club - Paisley Colour - I feel I need a view. There’s also a possibility that the federations will at some point will ask for views or issue guidance, so that’s another reason to get my thoughts in order. The intent here, then, is as much as anything to test my own thought process.
Before I start, there are a couple of caveats:
At the time of writing, the relationship between two presenters on the breakfast television box appears to be overshadowing war in Europe, a cost of living crisis, political corruption and incompetence as well as the supposed existential threat from AI itself. The impact of AI on photography seems slightly niche against these wider concerns.
Part of this will look at the impact on club photography. This is a topic about which you may know nothing and care less - that’s fine. This is a niche within a niche.
I will state at the outset, however, that I think some of the more outlandish warnings about the end of humanity from supposed tech experts seem to be more aimed at promoting their own name and brand rather than promoting informed discussion. There are issues with the untrammelled use of AI, but that’s a separate conversation…
I’m also sure that, as with Photoshop and digital, the debate will continue in both endless and pointless fashion, providing content for the letters page of Amateur Photographer for at least the next decade or so. In the meantime the sensible amongst us should simply pick up our cameras and get out to take some pictures.
TLDR version; photography will probably be fine.
AI Tools in Photography
Before opining on AI tools, it’s probably useful to map out the current landscape. In discussions with other club members and our club committee I’ve broken these tools down into three categories:
AI tools to facilitate editing. These tools don’t make any actual changes to the image, but rather make editing tasks faster, easier and more accurate. The obvious example would be AI masking tools in Photoshop, Lightroom and similar products. They use quite sophisticated algorithms and, for many, will save hours of work. However, any subsequent changes are entirely driven by the photographer.
At the other end of the spectrum you have the AI image generation tools such as DALL-E and Midjourney which generate complete images from text prompts. They’ve also generated a fair amount of headlines and press coverage. However, it’s pretty clear that while these tools are interesting in some quarters, and can create photorealistic images, they are digital art rather than photography. Making a clear distinction is a pretty straightforward task.
In the midground there are tools that are used to make changes to existing images as part of the editing process. The thin edge of the wedge is the raft of AI driven noise reduction and sharpening tools that create or change texture within an image to simulate sharpness. At the more extreme end of the scale are tools that relight a scene, turning it from summer to winter, or create components within an image from image analysis or textual prompts.
For many, and for the organisers of photographic competitions, it is this middle layer of tools that is particularly problematic. For some, however, these tools are undoubtedly a boon.
Not All Photographers are Competition Photographers
Any discussion of photographic tools and techniques needs to recognise that photography is a very broad church indeed. For some, the photograph is simply a product used in other activities; advertising, promotions, graphic design would all be decent examples. For these players, the option of being able to source cheap, AI generated content is attractive, particularly if margins are tight. Likewise, for some commercial photographers looking for royalty free textures or backgrounds, AI content is a valid option.
For more casual users, there’s also an appeal for the Instagram generation and any aspiring influencers. However, it has to be said that there is a large block in this constituency that abandoned any attempt to provide documentary content a long time back, and displays of intelligence, artificial or otherwise, are increasingly rare.
Will this mean that the opportunities for those wanting to make a career out of photography will diminish? Absolutely, but the prospect of the general public manning the barricades in defence of jobbing photographers and copy writers seems slim, particularly if the adoption of these tools means that they get their goods and services at a marginally lower price.
Professional Photography is Here to Stay
Does this mean that the role of the professional photographer will disappear? Absolutely not. Those proclaiming that AI means the end of photography as we know it are slightly missing the point. People buy photographs for a myriad of reasons, not least because a photograph can capture a memory or an experience or an emotion in a way no other medium can. A couple of examples to illustrate the point:
My youngest daughter is about to graduate from St Andrews University. We will, of course, order the official graduation photographs (as well as grabbing a couple of our own) as a record of a moment that makes us proud to be parents. I’m not sure any parent would opt for a cheaper option of an AI generated image of a student graduating - what you want is a record of the day and something that crystallises the memory.
For similar reasons, photography will always be in demand for weddings, children’s school photographs, family events, silver weddings and even, increasingly, funerals. There’s a deep-seated desire to record these landmark moments in our lives.
For the amateur as well, those photographs of loved ones and happy moments will always be something we cherish. For sure, AI may help polish the image, but not, I think, for many to the point of changing the reality.
The need for a record extends well beyond family moments. Sports photography, for example, can’t be passed to the AI to generate the imagery. The public wants to see a record of sporting excellent, not a vision of what it could have looked like. Similarly, the demand for documentary and news photography will remain.
For many, AI will support image making, but not supplant the demand for photographic imagery.
Trust and Photography
One of the comments I’ve seen more than once is along the lines of “I’ll never be able to trust a photograph again”. That strikes me as naïve; you never could trust a photograph. For as long as photography has existed it’s been used to distort reality; the Cottingley Fairies is a famous early example.
Ansel Adams (who I’m convinced remains so prominent in the public eye as much for his quotability as his photography) had at least a couple of relevant quotes:
The negative is the score, and the print the performance.
and
Dodging and burning are steps to take care of mistakes God made in establishing tonal relationships.
Adams is often quite playful and even humorous in some of his quotes, but it’s absolutely clear that he’s being very up front about the fact that what he presents in the final image is not necessarily a true representation of reality. Dorothea Lange (one of America’s most revered documentary photographers) put it more bluntly:
A documentary photograph is not a factual photograph.
Part of the problem is that we’ve grown up with the saying ‘The Camera Never Lies’ rattling around, but even that’s not quite right. The phrase has been attributed to the American actor Cesar Romero but the full quote conveys a slightly different message:
They say the camera never lies. It lies every day.
All of these quotes predate digital photography and Photoshop, let alone AI. The point is that even at it’s simplest, photography distorts reality. The framing of an idyllic scene of nature can, for example, exclude the industrial estate in the background; the choice of lens and viewpoint can change the relationship between foreground and background elements; and I can count on the fingers of one foot the number of people who have actually seen a 2 minute exposure.
Many have philosophised about photography (see Sontag or Barthes if you must), but the point is that the photographer has always presented a selective view of reality. The difference is that there are now more tools available to transform the original image and they are much, much easier to use. We have, effectively, democratised fakery in a range of mediums. If you want to take the dystopian view Orwell got there before most of us were born:
Every record has been destroyed or falsified, every book rewritten, every picture has been repainted, every statue and street building has been renamed, every date has been altered. And the process is continuing day by day and minute by minute. History has stopped. Nothing exists except an endless present in which the Party is always right.
- George Orwell, 1984
Orwell wrote this as satire; there’s a view that many politicians have adopted it as a playbook.
Amateur Photography: a Personal View
My own view is that amateur photography will be fine. For me, the final image is not the be all and end all and the process and experience is as much a part of the enjoyment as the end result. I have a preference for landscapes, but photography is as much about my enjoyment of the landscape as it is about the final image. It gives me a reason to pause, look, appreciate and enjoy my surroundings before pressing the shutter. Regardless of technology, that aspect won’t change.
The same is true for other genres, be it an appreciation of nature’s flora and fauna, the physical commitment and agility of sportsmen and sportswomen or capturing memories and moments. We’ll be fine.
The craft is also important. Despite the leaps forward with technology, there’s still an active community of the dedicated and dilettantes who, despite its limitations, continue down the film and chemicals route. Photography didn’t wipe out drawing and painting in the same way that the television box hasn’t eliminated the pleasure of reading. There will remain an active community who simply enjoy the photographic process, regardless of genre.
It’s also the case that for some there is satisfaction in the image making process. For these people the new raft of tools will be a blessing. Provided that they are prepared to be honest and upfront about their use of editing tools (and many are rightly proud of their abilities), does it matter?
Photography, as mentioned at the outset, is a broad church. For the foreseeable future, at least, there’s a place for commercial photography as well as those who take pleasure from the craft and experience of a very popular creative pastime.
Whither Club Photography?
Which brings us finally to club photography where, I think, a significant proportion of the angst and anger is on display. Firstly, a couple of positives:
Clubs, I think, will be fine. Membership will wax and wane as it has always done, but so long as there is an active constituency of people interested in photography as a hobby, then the clubs will have a role. Syllabuses will continue to offer insight, example and advice to those wanting to learn.
We also shouldn’t underestimate the important social function that clubs provide. Spending time with like minded individuals with a common interest is technology agnostic.
The real issue is around club competition, which is only a part of any club’s activity in any event. Competitions generally stipulate that all components of an image must be the work of the author; this is written into most sets of rules with varying degrees of detail and pedantry. Thus far the biggest threat has been plagiarism, which arguably requires effort and carries a non-negligible risk of detection. (It’s rare, but the fact that there are instances that come to light demonstrates that checks exist).
For images that are entirely generated by AI, there’s a direct parallel to be drawn with plagiarism. It may not be clear who ‘owns’ an image (there’s some ongoing debate as you can see here), but it is clear that it’s not entirely the work of the text box operator. Devising a clear policy to disallow these images should be straightforward. (Detection may be harder, but the problem is little different from that of plagiarism).
I mentioned earlier, that it’s the middle ground that presents the problem. There are a few factors at play here:
There is a spectrum of AI editing tools, from sharpening to AI generated fill. Their use, if done well, is also largely undetectable. There are questions then about where to draw the line (different folk will have different views) and, crucially, how to detect any breaches of the rules. Demanding photographers shoot in Raw and submit Raw images for scrutiny seems guaranteed to suck the joy out of competition even if you could find people willing to do the scrutineering.
The other main issue is that competition organisers have already enthusiastically opened the door to the use of AI tools already, through the wholehearted adoption of AI sharpening and noise reduction tools. Whether this was through ignorance, wilful ignorance or informed intent is debatable, but it’s not going to be reversed anytime soon. We’re now in a debate as to where we draw the line, if we even can. Put bluntly, you can’t police what you can’t detect.
There is also the issue that the line between some forms of creative photography and digital art has become very blurred. In this instance the challenge is not that the AI generated image is realistic, it’s that the ‘photograph’ is so obviously fake. It may be an unfair observation, but some of those who are clutching their pearls the tightest are those who have invested time and reputation honing this set of skills. AI makes what could have taken days quite straightforward, which is understandably distressing some.
(As an aside I recently had a conversation with a committee member at a club down south. He remarked that they no longer bother with external competitions as “it’s all elephants balancing on beachballs, flower ballerinas and over-processed portraits of grizzled old men”. I’m not sure I’d completely agree, but it was an interesting perspective).
It’s clear that those organising the larger national and international competitions face a challenge. Competitions with budgets, prize funds and corporate sponsors will probably go down the scrutineering route (at least for winners) if they’ve not already done so.
Salons and the photographic federations lack the budget and resource, so will need to find new ways to adapt. I think it is entirely possible they will end up taking their lead from their member clubs, rather than assuming that they can tackle the issue from the centre.
Trust the Membership
Ultimately clubs will need to trust their membership. These tools are here to stay and, even if we wanted to, we’re not getting the AI baby back up the birth canal. Competition secretaries and club committees will need to help themselves and their members by delivering good guidance. Personally, I’d favour principles rather than proscriptive sets of ‘shalts’ and ‘shalt nots’. The difficulty with going down the route of a proscriptive set of rules is that the technology is evolving rapidly, meaning that the ruleset will need to evolve at a similar pace.
Instead, we will need to be more thoughtful about the outcomes of processing rather than the mechanics of the edit; focusing on the ends rather than the means. We’ll also need to assume that our members (who, after all, are in the main functioning, adult human beings) have the judgement and sense to understand the guidance and wit to ask when it’s not clear.
We’re helped by the fact that for most, if not all, the aim of entering a competition is to garner some feedback on their images, technique and style. It’s as much an opportunity to learn and improve as it is to gain recognition and kudos. ‘Cheating’ teaches us very little, and I think most of us understand that point.
We also know our members. There are a couple in our club, for example, who if they started submitting beautifully composed ICM images you’d suspect an altered state of mind before malpractice. You’d at least ask the question…
As a final point, I think this influences what we ask of our judges, with less emphasis on ranking and scores and more on feedback and critique. Our club, as have many, has already moved in this direction; more I suspect will follow.
Finally, I’ll end with a plea for those entering competitions to put themselves in the shoes of others for a moment. Most people are members of a club because they care about their photography and want to improve. They may have a different style and approach to you, but they probably care as much about the integrity of their own work and competitions as anyone else.
There are very few bad actors in the club community. Technology doesn’t make these good people into bad citizens - ultimately tolerance and curiosity will be more effective at policing behaviour than any set of arbitrary rules.
It’s a hobby. Enjoy your photography; improve, appreciate and have fun! We’ll be fine…
