When you ask Siri, Alexa, or Google a question, you have to say the trigger phrase, “Hey Siri”, “Alexa”, or “OK Google”. When you want to follow-up, you have to repeat that trigger phrase.
But that is about to change.
Amazon has rolled out a beta of something it calls Follow-up Mode. From this Amazon Help page:
When you turn on Follow-Up Mode, you can make more requests without repeating the wake word. Follow-up mode is available on all hands-free Alexa-enabled devices. The blue indicator light on the Echo device remains on for a few seconds, letting you know that Alexa is active and ready for your next request.
The idea is, after you say “Alexa”, Alexa keeps listening for a follow-up query until it times out. Notably, follow-up mode does not kick in when:
- Audio is playing. For example, the device is playing music, books, or you’re on a call through the device.
- You end the conversation with Alexa. For example, you can say, “stop,” “cancel,” “go to sleep,” or “thank you.”
- Alexa is not confident you’re speaking to her. For example, if she detects that speech was background noise or that the intent of the speech was not clear.
I’d think this would be relatively simple for Apple to implement for Siri. There’s certainly value when you want to do a sequence of things. But I also think it’d be a big help for Siri’s context analysis, making it more likely that two queries in a row are connected in some way, like so:
- Hey Siri, what time is my next meeting?
- What time is the one after that?
Or:
- Hey Siri, what song is this?
- What album is it from?
You get the idea. Conversational context is a bit of a holy grail. As is, all three (Siri, Alexa, Google) are still infants, still learning the most rudimentary rules of conversational memory.
I’ve done follow-up conversation with Siri many times using HomeKit devices.
I’ll say “Hey Siri – turn on the family room light”
(Siri will turn it on)
wait ~10 seconds
I’ll then say “Turn off the light” or “dim the light to 50%” without saying “Hey Siri” first and it works.
Not sure if anyone remembers, but the MotoX several years ago had a hands free mode that would automatically kick in when you were driving and it had this sort of conversational feel where it would remember context. If I remember correctly, it was limited to interacting with phone calls and messages, but it was really good and felt really fluid. It will be interesting to see how Alexa implements this.
Hasn’t this existed since the debut of Siri? And it is in “Hey Siri” as well — I can often say “Hey Siri, what’s the weather today?” then follow up with “What about tomorrow?” That said, it doesn’t work with every query, and more often than not, it doesn’t work at all. Perhaps the narrative here is that Apple hasn’t been able to implement it in a good way?
Simon Says.
What I’ve always wanted from Siri is…
“Hey Siri, remind me to walk Kylie at 5pm.”
“Okay, I’ve set a reminder for 5pm.”
“Kylie is spelled K-Y-L-I-E.”
“Noted.”
But in this case, actually, Siri has figured out it’s Kylie, so I don’t have to remind her about my dog’s name. But so many times it misspells (often badly) what I said. I would love to be able to follow up with a spelling correction.
Oh, man. Err. Well enjoy it.
I’m pretty sure this would absolutely suck.
Might be good…if it works.
I got tired of watching tv and Alexa being triggered so I finally unplugged it and threw it in the closet. Never could figure out why, listening to the app nothing ever sounded remotely like Alexa. But it would be hell for it to be triggered this way and then to continue trying to converse with the tv. If I were an armed man I might end up pulling an Elvis and shooting it.