I don’t think many people really understand how LLMs work. Why they manage to come across as if they “know” things, even though they’re “just” a model of language, not of knowledge (though this book also raises the question of whether the two can really be separated at all). Why being wrong is baked into how they work.
I read this book cover to cover during two rather long waits at the doctor’s. It’s a small paperback, but one that gave me real insight in several places.