AI firms are buying old books for slop-free training data
SMRTR summary
AI companies are quietly buying old physical books — printed before 2022 — to use as clean, human-written training data, avoiding the growing flood of AI-generated text online. Data broker ISBNdb sells bulk books to labs, often destroying them by slicing off spines to scan pages, while keeping buyers secret under strict NDAs.
SMRTR provides this summary for quick context. The original article belongs to Daily.dev.
Read the original article