As you probably know, the new Apple iPod shuffle can speak up the name of your songs, artists, and playlists. I initially thought it was a new text-to-speech chip. It is not:
As you can see in this flowchart, the trick happens in your computer. Apple patented the process a while back, and it's going to implement the mechanism in the next revision of iTunes. iTunes 8.1 will use the text-to-speech capabilities of your computer to generate sound bites, embedding the name of the artist and the song in every tune.
This explains why the voices are different depending on the platform you sync your iPod shuffle with: iTunes 8.1 uses the text-to-speech engines available on each operating system with different settings. The only thing that the iPod does is playback those sound bites when you click on the central button, located in the headphones cable (sadly.) [9 to 5]