Generative AI does not respond to intention. It responds to language. The quality, clarity, and
precision of what a user types determines the quality, clarity, and usefulness of what the tool
returns. This paper demonstrates that mechanism through original research conducted by the
author using two leading generative AI tools and two levels of prose literacy in the prompts. The
findings are reproducible, immediate, and carry significant implications for every educator who
wants to understand why prose literacy is not simply related to AI capability but is its
foundational prerequisite.