I don’t think locking the content of rare books away in the hands of corporations who only give us access to tools trained on the books, and not the actual book, is a good path forward.
This doesn’t incentivize them to be good stewards of this data and making anything in the public domain available. It incentivizes less access to the source material, having to blindly trust their tools, and is effectively automating plagiarism.
Also, why would they ever share their collection?
Book scans, secreted away, are worthless to the public.
They could use them as training data, without providing access the actual books.
At least we would all benefit from the books this way, so long as legal nonsense keeps the scans unavailable to the public.
I don’t think locking the content of rare books away in the hands of corporations who only give us access to tools trained on the books, and not the actual book, is a good path forward.
This doesn’t incentivize them to be good stewards of this data and making anything in the public domain available. It incentivizes less access to the source material, having to blindly trust their tools, and is effectively automating plagiarism.