Imagine waking up tomorrow with the solution to the Voynich Manuscript.

[quimqu]

Oddly enough, LLM's have read a lot, keep on reading, and are likely to catch on.

This is an interesting point. In my opinion, LLM can be used as a huge knowledge summarizer, sort of Google sumarizer, and they can be used as idea creators (cautiously) (for example Terence Tao is using LLM for pushing his mathematical experimentation).

The bad point is, that they "keep on reading" and that now, a big part of the internet is created by LLM, with their hallucinations, and that part is increasing a lot. So what will they learn in a future?