Why is default zero initialisation useful

[KeAiMianYang]

I mainly come from higher level, garbage collected languages, not from C. So I am not talking compared to initialised behaviour in C.

Even after reading about ZII and making the zero value useful, I do not understand in practice why this is valuable — especially for nil pointers where it seem to me that the ability to represent non-nillable pointers is more valuable.

What kind of patterns or code examples can I check to see the point of ZII?

Partially related, how painful is the absence of non-nillable pointer in practice? Coming from Java and C# I cannot imagine being fine without the ability to represent them.