I wish there were some kind of tool that would do a diff between two audio files so I could string together all the ums and ers and I means I've cut out and send them to the person that others praise for being so eloquent and fluent.