How Amazon ejected AI-written e-books from its bestseller lists

Topic

New Reply

PUBLIC DEFENDER By Brian Livingston The giant online retailer, Amazon.com, faced a problem. Hackers were using chatbots to create fake e-books — mostl
[See the full post at: How Amazon ejected AI-written e-books from its bestseller lists]

9 users thanked author for this post.

Myst, fisher75, Steve S., ibe98765, Norio, Izzienoel, rc primak, DLivesInTexas, abbodi86

Reply | Quote

Replies

Wait you love me is still a better love story than Twilight.

Reply | Quote

This brings up something I’m quite concerned with: I’ve contended that the old adage “garbage in, garbage out” is one of the significant long-term risks of AI.

Consider that LLMs are being trained by scrapeing the internet of its massive collection of human language. Then consider that the output of AI chatbots are quickly beginning to flood the internet – and will no doubt strongly accelerate in the future as websites, news articles, books, papers, blogs, social media, ad infinitum, use AI to generate content. AI will then continue scrapeing the internet, sucking up an ever increasing amount of AI generated content for its ongoing training. This is a GIGO feedback loop if ever there was one.

Will this feedback loop lead to an unstable degredation of human language and knowledge or will we be able to create controls on that feedback loop, keeping it from spinning out of control? Having witnessed the negative effects of social media on our society, I’m not so optimistic.

Win10 Pro x64 22H2, Win10 Home 22H2, Linux Mint + a cat with 'tortitude'.

2 users thanked author for this post.

LH, Myst

Reply | Quote