Three ways to choose what to build:

> solve a problem one personally feels;
> ask what will become necessary in the future;
> enter a validated category with something proven, better, and new.