Journeys in AI photography - thoughts on Midjourney.

AI generated portrait photograph of a woman standing in a misty room under neon lights. Made by Stefan jakubowski using Midjourney V5 AI generated portrait of woman standing in front of white neon lights Photography made with AI of woman in aquarium generative AI photograph of woman in neon-lit apartment at night woman in fairground inside neon-lit house of mirrors. Mature woman wearing sunglasses with glowing green neon frames. Portrait photograph of blonde woman generated by artificial intelligence Old man in front of balloons, AI photography Woman in futuristic pink car driving through Tokyo

“What is “real”? How do you define “real”? If you’re talking about what you can feel, what you can smell, what you can taste and see, then “real” is simply electrical signals interpreted by your brain…” 

Morpheus, The Matrix, 1999.

I’ve been thinking about this quote a lot recently. 

In my previous blog post, I wrote about how I felt anxious about my own usefulness and future as a photographer after seeing the trend on Instagram of users creating high-quality painterly portraits of themselves using an AI-powered app at the end of 2022. I hadn’t had the chance to use AI in my own work up to that point; my only exposure so far had been trying out DALL-E earlier to create a few fun images of an imaginary sci-fi movie that some friends and I came up with in the pub one night. Whereas in the past I’d probably have used photoshop or Canva to knock up a fake and (hopefully) funny poster, I thought this was the perfect opportunity to try out generative AI. The tool was quick and I was pleased with the results but overall I didn’t think of it as anything that I’d be able to incorporate it into my own photography work and didn’t use it again after that.

Then I heard about the work of Tim Tadder, a commercial photographer who made the news when he announced he’d no longer be shooting real images but instead creating all of his work going forward using the generative AI tool Midjourney. I began following his Instagram page and was blown away by how realistic the lavish and colourful images he was able to create using only a few text prompts were. 


One post in particular really piqued my curiosity and made me want to try it out. He described a shoot he’d always wanted to do in Paris featuring models standing inside giant colourful bubbles, but for various reasons wasn’t able to make it happen. But using Midjourney he was able to realise his concept and bring the images he’d dreamed of to life, with suitably stunning results.

While being incredibly impressed by the quality and realism of the work being produced by Tim and other AI art creators, I was still anxious about whether this might mean that real photography, and photographers like myself, might be in danger of being replaced. And so, I decided, it was time to try using generative AI for myself; to see what it could do, what it can’t, and whether my fears for my career were justified or not. 

And I was not prepared for the impact it would have on me. 

I didn’t have any expectations, or even any specific ideas for images when I first signed up. I simply wanted to try the process of writing prompts to produce images, and test out the results and learn about how it all worked. I’d only just done a collaborative shoot with a model a couple of days previously, and I certainly didn’t have any intention of following Tim Tadder’s lead in abandoning in-person shoots. But as soon as I saw the first few images from my prompts I knew that this was going to be an absolute game-changer. Some of the images I’ve been able to create have looked so real it’s astonishing - you can see a selection of some of my creations in the gallery above. 

I’d be lying if I said I haven’t thought about hanging up the camera when it comes to creative/concept-based shoots after seeing some of the photorealistic imagery Midjourney has enabled me to produce. But, here’s why I think, despite everything I’ve said so far, that generative AI does not spell the end of photography, but rather the start of a new chapter for it.

It takes more work than you might think.

Sometimes you get a prompt right first time and get exactly what you want, but more often than not to get the imagery you want means writing and re-writing prompts, rearranging them, using other photos as prompts or running the same prompt several times. It still takes work to produce great images even with AI’s help. 

We have to honest and say that some photography work might be replaced by what tools like Midjourney can do - stock photos being a very likely early casualty - but I feel that people who have a technical understanding of photography, lighting and image creation in general will have a massive advantage in knowing how to describe things in a way that gets professional results. Much like knowing the settings on a camera for a given situation, our experience and knowledge will still be of huge value even in the age of AI.

Creation is still important but curation is key.

What I’ve found with the process above of re-writing and re-running prompts, making variations etc, is that often, when you’re creating sets of images based on a specific theme or idea, you can very quickly end up with a large selection of images. This is where, much like on a photoshoot, the key to a great set of images is being able to choose the ones that you feel will make the biggest impact and deliver on what you wanted from your concept. Generative AI outputs image data based on what you type in - it’s still you that has to decide whether what it gives you is ‘right’ or ‘good’.

It’s not either/or.

Real-world, in-person photography isn’t going anywhere. There is still a need for it and it will still be there for those who want to do it. AI if anything might help you make your real shoots better through being able to produce more sophisticated and tailored moodboards, test different colour combinations, styling, makeup and lighting scenarios. Or even just give you new ideas you might not have thought of. 

The invention of the camera didn’t replace painting, drawing, or sculpture. Artists still work in all these mediums and audiences still enjoy them. The camera simply added something different. Portrait mode on smartphones hasn’t replace portrait photography. In the same way, AI will be another option to add to the many ways that creators can create.

Ideas still come from you.

Having the highest-quality paper and the best pen, or a nicer laptop, doesn’t make any difference to the quality of your stories if you’re a writer. Similarly, a tool like Midjourney is a visual blank page - it’s still up to you, the photographer, to decide what you want to fill it with.

‘Real’ or not, there’ll always be an audience for good-quality content.

There will always be an appetite for new music, writing, films, games or photography. So whatever tools you decide to create with, if you’re making good work, it will always have impact, and I believe, value. 

In conclusion, for photographers (or any other creatives for that matter) anxious about how AI will impact your work going forward, my advice is simple - try it for yourself. You might be surprised at what you think of it, and how, rather than being something that might replace your work, it could be something that could help enhance it.

I’m sure I’ll have plenty more thoughts on Midjourney and other generative AI tools going forward - I’m already seeing many of the creators I follow on Instagram experimenting with AI text-to-video services and it’s clear this is going to be the next frontier for the technology so I may well give that a try next.

Using Format