.. yeah a startrek set would be cool...
well in fact i think you guys should make the whole voice recognition software as configurable as possible.. so users can generate their own 'user interfaces' for voice recognition (empeg responds over speakers with confirmation/wishing me a good day/asking me to repeat a command if it didnt understood etc)...

i dont think you should only make playlist names trainable... what about volume controls for example.. i like the idea of beeing able to tell my empeg to lower the volume or so...

will the recognition software be able to understand my command even though the music is playing .. ie.. subract the empeg output from the audio signal the micro receives ?