Comment by blendergeek

I think it needs to be continuous speech-to-text as well.