The audio desktop is the Emacsvox working environment. It is not a separate program or a spoken copy of a graphical desktop. Emacs keeps each document, application, and result in a buffer; its modes understand the kind of information in that buffer; and Emacsvox presents that context through speech, voice changes, sounds, and structured navigation.
You can begin without learning a new desktop model. Open a file or application normally, use C-e m to hear the current buffer and mode, and press C-h m to discover what that buffer can do. The rest of this chapter explains why the same approach works for text, Web pages, mail, shells, source code, and many other tasks.