Maybe you missed this discussion https://news.ycombinator.com/item?id=36400053 https://news.ycombinator.com/item?id=36400053 it seems OpenAI is aware of their s

Maybe you missed this discussion https://news.ycombinator.com/item?id=36400053 https://news.ycombinator.com/item?id=36400053 it seems OpenAI is aware of their software outputs copyrighted stuff so they attempted some quick fix filter. So the fact it will not ouput the book for us when we ask does not prove that the AI does not have memorized big chunks of it, it might just be some "safety" filter involved and you need soem simple trick to get around it.