notes for Jimmy:
    * if spoken word is empty, restart recognition
    * add ability to get hints
    * music on/off
    * consider changing the language buttons
    * code refactor part 1