Then it’s gonna run speech recognition on what you’ve requested it to do. Speech recognition goes from audio to textual content; it’s transcribing what it's that you’ve requested for. So it says, hopefully, "Play me some music". Though speech recognition methods usually are not excellent, and so they make some errors, we hope that more often than not although it will get your request correct sufficient when it transcribes i "" Invoice Hinton/Second/Getty ...