The AI industry tells its story as an engineering feat — build the model, scale the compute, ship the API. The publishing industry lived the same plot five centuries ago. Gutenberg's press collapsed reproduction costs by orders of magnitude, and within two generations it was commodity hardware. What endured was the publishing house: the institution that decided what to print, how to organize it, and how to reach readers. This collection traces the structural parallel between publishing and AI — not as metaphor, but as an industry-structure comparison. Model training is entering its commodity phase. The competitive axis is shifting to data sourcing, editorial judgment, content packaging, and distribution. These are publishing questions. They have been publishing questions for five hundred years.