Slop is not a technical failure. Most of it is technically clean: correct anatomy, pleasant lighting, no artefacts. What it lacks is a decision. Nobody chose the lens, nobody chose what to leave out, and nobody was willing to make the frame worse in one way to make it better in another. The audience reads that absence instantly, even when they cannot name it.
The commercial cost is specific. Slop trains an audience to skip a format. Once a feed has taught someone that a certain gloss means nothing worth reading, every subsequent piece in that register pays the tax, including the good one.
The fix is not less generation. It is a person with taste holding a veto, and a brief that states what the piece is allowed to be bad at.
How do you tell AI slop from good generative work?
Ask what the piece is refusing to do. Good work makes a choice you could argue with: a strange crop, an ugly colour, a beat held too long. Slop makes no choice at all, which is why every piece of it looks like every other piece.
Does using AI automatically produce slop?
No. Slop is what you get when the tool is the whole process. When generation is one stage inside a pipeline with gates, a register and an editor, the output looks like the studio rather than like the model.