When I speak on my Apple watch - the text to speech is quite fast. (To me this seems like the harder Computer Science problem, and my understanding is that it is going over the wire - back to Apple - converted, and then sent back as text).

Then the subsequent action "Tell John Smith I'm on the bus" - launching an SMS with this content, seems quite slow.

This seems like the wrong way around - since I would have thought that the speech-to-text and running it over the wire was much harder to engineer.

My question is: Why is my Apple Watch Siri voice-to-text fast, but taking subsequent action is slow?

  • Taking action requires "voice-to-text" and then determine what action to take. So there's additional step. "Tell" -> open Messages, "John Smiths" lookup in Contacts, "I'm on the buss" -> voice-to-text fill SMS. – Mateusz Szlosek Aug 25 '15 at 11:22
  • 1
    Yeah but that doesn't involve taking an analogue signal around the world, pattern matching it against millions of voices, parsing it, and bringing that back in sub-second time. It is more of a localised text parsing problem. An order of magnitude simpler IMHO. – hawkeye Aug 25 '15 at 11:29

You must log in to answer this question.

Browse other questions tagged .