AGI could be possible if a new breakthrough is made. Currently LLMs are just pretty good text predictor, and any intelligence exhibited by them is because they are trained on texts exhibiting intelligence (written by humans) . Make a large enough model, and it will seem like an intelligent being.
Probably some children’s book I can’t remember. The earliest book I remember reading by myself would be one of this series (in Polish) https://en.wikipedia.org/wiki/Tomek_in_the_Land_of_Kangaroos?wprov=sfla1
This one was also a popular children’s book, but I think it was read to me by my mother https://en.wikipedia.org/wiki/Akademia_pana_Kleksa_(novel)?wprov=sfla1