Stop Defending Human Value With Benchmark Comparisons to AI

The common defense of human worth in the AI era—that humans still outperform models on certain tasks, or produce subtler output—rests on a shrinking capability gap. Tying human value to a moving benchmark is fragile reasoning; the goalpost shifts with every model release. The stronger claim is unconditional: humans are valuable, full stop, with no qualifier about output quality required.

The piece then pivots to what “quality” in creative work actually means, distinguishing intent from material form. Creation is normally the slow distillation of intent into form, with the two inseparable. Generative AI breaks this link by producing substantial form from minimal intent—a user with a vague mental model can still ship a polished artifact. The author reframes “AI slop” as form without discernible intent, noting humans can produce it too; AI has just lowered the barrier. One cited suggestion: if you’re going to send an LLM-written email, send the prompt instead, since at least the prompt reveals what you meant.