Photographic Slop

Everyone is talking about AI slop. Nobody is talking about the human-made version that has been flooding the industry for decades.

You know what AI slop looks like. The uncanny smoothness. The slightly wrong hands. The generic beauty that scans as "professional" without carrying a single original thought. It fills feeds, it fills presentations, it fills pitch decks. It is the visual equivalent of elevator music: technically competent, emotionally empty, produced at scale.

The industry is furious about it. Photographers are furious. Art directors are furious. Creative directors are furious. Everyone agrees that AI slop is a plague on visual culture.

Here is the problem with that outrage. Most of the brands complaining about AI slop have been commissioning the human equivalent for years. It just cost more to make.

The slop you aren't talking about

Open the last five campaign shoots from most mid-to-large fashion brands. Strip the logos. Shuffle the images. Now try to identify which campaign belongs to which brand.

For the majority, you can't. Because the images were not made to be distinctive. They were made to be appropriate.

Appropriate lighting. Appropriate casting. Appropriate location. Appropriate styling. Appropriate retouching. Every decision calibrated against the same set of references, approved by the same chain of stakeholders, produced by the same rotating cast of photographers, stylists, and post-production houses.

That is photographic slop. It is beautiful, expensive, and completely interchangeable.

How it gets made

It starts with a mood board. The campaign, the editorial, the brand shoot: almost all of them begin with someone collecting images made by other people and presenting them as a "direction."

Those images were themselves produced from mood boards that referenced earlier images. The photographer interprets the references. The retoucher matches the references. The final image looks like the references, because that is what was asked for.

Each cycle narrows the range. Each brief that begins with "something like this" produces an output that becomes someone else's reference next season. The result isn't bad work. It is averaged work. Work that has been filtered through so many rounds of consensus that every sharp edge, every odd decision, every moment of genuine surprise has been sanded smooth.

That sanding process isn't an accident. It is the system working exactly as designed.

Why the outrage is misplaced

The reason the industry is so upset about AI slop isn't because AI produces bad work. It is because AI produces similar work, faster and cheaper, and that makes the existing process visible.

When a creative director assembles a mood board, briefs a photographer, art directs a shoot, and approves a set of images that closely follow the references they started with, that process is hidden behind production infrastructure and the mystique of "creative vision." It takes weeks. It costs tens of thousands. It employs dozens of people. The expense makes it feel like a creative act.

When a machine produces a similar result in thirty seconds from a two-line prompt, something becomes visible. Human photography contains real value: creative partnership, real-time decisions on set, the relationship between photographer and subject. But when the brief, the references, and the approval chain have already determined the output before anyone picks up a camera, that value is being spent on reproducing conventions, not creating distinction.

What actually survives

There is a version of human image-making that AI can't replicate. But it isn't the version most brands are buying.

The version that survives is the image that contains a decision no reference board would have suggested. The composition that is wrong in a way that makes you look twice. The lighting choice that violates the category convention and, in doing so, becomes the only image you remember from the entire season.

That kind of decision doesn't come from a mood board. It doesn't come from a training dataset. It comes from someone with enough experience and conviction to know where the convention should be interrupted, and enough authority to defend that interruption through the approval process.

Some photographers produce that. They are the ones whose work you can identify before you see a credit. They are also, not coincidentally, the ones least worried about AI.

Most of the industry isn't set up to produce that. Most of the industry is set up to produce consensus, and consensus is slop whether a human or a machine makes it.