The real problem with Siri is not that it (often) goes down, but rather how Siri informs you that it is down — which it does after you have already asked it to do something.
Apple, and Siri users, certainly want the service to maintain 100% availability, but with services that rely on so many working parts — this is just an unrealistic expectation. So when Siri fails it should do so in the least intrusive way — it just so happens that’s not the way it fails right now.
Typical scenario: Driving home and want to text my wife to let her know that I will be there in 20 minutes. Activate Siri and ask her to message my Wife. Siri comes back to say: something, something, I don’t work.
Siri really should be smarter than this. Siri should tell me from the moment I activate it whether, or not, what I am about to do is going to work. How hard is that?
– Activate Siri.
– Siri checks servers.
– Siri responds: “Sorry, I’m drunk as a skunk right now. I’ll sober up soon.”
That’s really not that hard and offers a far less frustrating experience than the current solution of leading on the user, in hopes that — what — it will be back working by the time I am done speaking?