Amazon Destroys Rare Books to Train AI Models

Amazon is acquiring out-of-print and rare books—including first editions and limited runs—then pulping them for AI training data. This erases irreplaceable cultural artifacts for marginal model improvements. The practice reflects a collision between tech's data extraction logic and cultural preservation: physical destruction, unlike digitization, is irreversible. Disposal costs less than proper archiving, making destruction economically rational. Other tech companies will likely follow once legal and reputational costs prove manageable.