LLMs are creating a huge sanitation problem that will probably never be solved
Taught a class on AI this summer. For the first two writing assignments the students were given a subject and required to use an LLM of their choice (mostly ChatGPT) to create the first draft. Then rewrite the paper.
Both versions were submitted to complete the assignment. Then the papers were posted on a discussion board and each student was required to review 3 other papers.
The rewritten paper was consistently better than the LLM version, both in content, organization and readability.
I tried to drive home the point that "You write to be read.". If you generate and publish LLM - pretty soon no one will read you.
Frankly, the internet has been a huge cesspool for quite awhile now. You could deal with it by paging thru Google results and sorting out the clickbait. But even that technique was getting harder.
Did you see "When AI Is Trained on AI-Generated Data, Strange Things Start to Happen" by Maggie Harrison in Futurism?