When producing costs nothing, your acceptance criteria are worth noth…
Three verdicts in one week on work made with a machine: one accepted without debate, one still disputed, one awarded then flagged as artificial by a detector. What separates them, what your grid actually measures, and 3 steps to rewrite it.
Three verdicts landed in the same week, on three pieces of work made with a machine. The first collapsed a maths problem that had stood open since 1939, and nobody contested it. The second won twenty five thousand dollars in a well known contest, and the accusation that it was churned out by a model is still unresolved. The third received a literary prize, and when its author ran his own hand-written article through a detector, the tool declared it 82 % artificial. Three opposite outcomes, one factor separating them. It has nothing to do with the quality of the work. Why it matters (PO, designer, design lead running a component catalogue): every line in your grid that rewards visible effort now favours whoever did the least work. This week's move: take a grid you actually use and rewrite…