“Speech Generation and Sound Understanding in The Era of Large Language Models” David Harwath