Good luck inventing something new and better without first understanding what already exists. In my experience it's much more efficient to master what already exists, first, noting down any novel ideas as they come to you, and then move on to implementing those novel ideas.