Amazon has created a new rule limiting the number of books that authors can self-publish on its site to three a day, after an influx of suspected AI-generated material was listed for sale in recent months.
The company announced the new limitations in a post on its Kindle Direct Publishing (KDP) forum on Monday. “While we have not seen a spike in our publishing numbers, in order to help protect against abuse, we are lowering the volume limits we have in place on new title creations,” read the statement. KDP allows authors to self-publish their books and list them for sale on Amazon’s site.
Amazon told the Guardian that the limit is set at three titles, though this number may be adjusted “if needed”. The company confirmed that there was previously no limit to the number of books authors could list a day.
What’s this easy fix then? Just a lower number? That will just mean more publishers.
AI detection tools don’t work, and humans aren’t much better, unless they’re subject experts. How do we stop AI books?
You don’t. There are plenty of legitimate cases for using AI (translation, spell check, summaries, etc.). The solution should be to use AI to categorize all the books and their content, something in the style of Steam-tags, but going even deeper into what is actually in the book and done by machine, as nobody has time to read all those books.
This is kind of the biggest surprise in this recent rise of AI: We got to generation before categorization took of. Even so a lot of the AI work in the last 10 years has been focused on categorizing, nobody has build anything with it, at least not user accessible. Everything is still running on human made categories. The only time you see categorization in action is in the recommendation algorithms of Youtube or TikTok, but the categories that drive them are hidden from the users.
I’m waiting for a self hosted organization ML that can find my desired data in the thousands of documents and pictures