Terrible images are going to save photography from AI – and this is why

Motion, blur, out-of-focus shots, harsh direct flash – some of the biggest trends in photography right now are the very things that I first learned to avoid as a beginner. But these so-called “bad” photographs that break all the rules aren’t just trending; these “terrible” photographs are what’s going to save photography from AI.

I don’t think it's a coincidence that the retro photo look is returning at the same time that computers are becoming capable of generating “perfect” images.

Generative AI is becoming increasingly accurate at creating photo-realistic graphics with fewer errors. With a training data set filled with sharp, high-resolution images, AI is becoming far more capable of spitting out images that look like they were taken by a computer that memorized all the photography tutorials.

Yet, at the same time, shoppers are spending hundreds of dollars on cameras that do less, not more. Cameras like the Fujifilm X Half that intentionally limit the features in a film mode are becoming best sellers. Disposable film cameras are making a comeback (along with digital dupes).

Lenses and filters are launching with imperfect image quality – and listing that imperfection as a selling point. Point-and-shoots are returning as AI increasingly automatically edits every photo snapped with a smartphone.

Yes, some of the rise in intentionally capturing imperfect images is due in part to the rise of retro tech in general. I can now buy actual records at the same store where I buy my groceries, and retro-styled cameras are among some of the most popular models.

But I think part of this rise in retro-styled, imperfect imagery has simpler roots: we’re growing increasingly tired of digital perfection. I’m craving authenticity and simplicity in an increasingly digitally altered world.

The photographs that intentionally break the rules and do it well are “terrible” only by the traditional definition of what makes a photograph “good”: a sharp subject, an interesting composition and good lighting.

The “terrible” images that are going to save photography have one major feature in common: emotion. The photographs that break the rules and break them well do so to create a photograph with far more emotional impact.

Photographs with motion blur put the whirlwind feel of how quickly time passes into a single image. Images with old-school colors and grain give memories a nostalgic feel, even if those memories are only from yesterday.

AI is becoming increasingly capable of creating images that look realistic – and even images that actually look like real people. But I don’t want perfect photographs.

I want photographs that remind me of that awkward smile when my kid lost a tooth. I want photographs of my daughter’s messy morning hair, of my son’s sand-covered toes at the beach. I want photographs that remind me of not just the memory, but the feeling that I was experiencing in that moment.

AI can improve – training has already reduced the number of times an AI will generate extra fingers and badly spelled text. But a computer can’t feel emotion – and that’s exactly what is going to save photography from AI.

You may also like…

Learn how to spot an AI-generated image. Or, browse the best film cameras or the best compact cameras.