Why is Anthropic destroying books?
The AI company apparently found destructively scanning ‘all the books in the world’ easier than dealing with copyright in its quest for training data
TL;DR
- Anthropic's 'Project Panama' involved destructively scanning books to create a high-quality dataset for training its AI model, Claude.
- The process involved vendors slicing book spines and discarding original copies after scanning.
- A lawsuit, Bartz v Anthropic PBC, highlighted this practice, resulting in a $1.5 billion out-of-court settlement with authors.
- The court ruled that using proprietary material to 'train' an LLM did not, in itself, constitute copyright infringement.
- The practice raises concerns about the future of books as cultural heritage and the role of human authors in an AI-dominated landscape.