Voice to text integration

Looks like there has been some talk about that here: