No clicks, no content: The unsustainable future of AI search

I've written a 125,000-word book a year before GPT-3 was a thing.

If this book came out today, in 2025, how would you know that the 420 pages are actually worth your time and not just a bunch of hallucinated LLM slop?

I've been wondering whether Wikipedia and libraries in 2030 will be in a better overall place quality-wise, or will just be overrun.

The last few times I looked for information on YouTube (by typing a keyword phrase or question instead of looking up a specific channel/creator), most of the top results were AI-narrated presentations. Some of those were filled with comments of people correcting obvious mistakes in the content (which as a layperson I would not have seen as mistakes)

Wikipedia would probably have to deal with its internal politics and frequent-contributor culture before 2030. Wikipedia pages are not too shabby, in expectation. The Wikipedia meta on the other hand is something much less pleasant.

This is music discovery for me now. Anything new by new artists with zero background/presence/indicator that they wrote the music I just pass on.

It seems like an unrelated issue to the topic at hand, though. I'd rather have difficult human interactions and trustworthy content VS no human interaction and hallucinated content.