Really? Like what deep questions?

For me LLMs are a step forward from previous ML approaches (some of them were and still are extremely useful anyway) and for sure are very useful to many people. They are very good at generating sensible output based on input created by millions of people, but very poor at creating anything really novel.

> Like what deep questions?

Is intelligence purely computational?

Is consciousness purely computational or at least an emergent property of it? If so do natural numbers have consciousness?