Apple's Intercom Feature Works on iPhone, iPad, HomePod, Apple Watch, AirPods, and CarPlay

[Alan Wynn]

Yeah I’m still unclear how some of this works. Can’t wait until we can test this out. Though I probably won’t be able to test out any feature that depends on the U1 chip as I have regular homepods and don’t plan on getting minis. Maybe one, we’ll see.

I expect that the next AppleTV will have a U1 chip and coupled with HomePods in the same room will provide much of the same functionality.

Also I assume all that voice data has to be sent to Apple then back to one’s home. It would be great if it could never leave the home.

I am not sure that it goes back to Apple (at least for connections that are only in the home). It would be cheaper and easier for them to have it be local traffic only (and would also eliminate some regulatory issues).

Of course even more cool and ideal would an option to go into a phone-like conversation where you can not only hear speaking real time but you can talk over each other, like the person is in the room with you, until you say “hey Siri end call” or something. And none of this data would be sent to Apple or a phone carrier ideally.

That makes the most sense for one-to-one conversations (something it may or may not do right away). Since these are not intended to be stored, I would expect that the traffic would all be encrypted with ephemeral keys to ensure privacy, but we will have to see.

 

[MacAddict1978]

Finally, a feature where Apple is a notch ahead of everyone in being able to broadcast to a phone. Which is also the first feature I'd disable if I had a family at home. Because we all love someone broadcasting through our phone during an 8 am meeting they can't find any clean underwear.

At least the Home Pod is starting to get some features that make it a bit more useful as a smart speaker aside from it's assistant being mentally incapacitated most of the time.

 

[H3LL5P4WN]

No we are reviving NuTone home intercoms, a staple of a suburban mansion up until smartphones became a thing.

Hell, even in park-model trailers.

 

[subjonas]

Finally, a feature where Apple is a notch ahead of everyone in being able to broadcast to a phone. Which is also the first feature I'd disable if I had a family at home. Because we all love someone broadcasting through our phone during an 8 am meeting they can't find any clean underwear.

Hahaha. True, I wonder if Apple foresaw this situation and put in a “knock knock” mode where during a certain time a certain homepod will only allow a notification sound that one can choose to respond to, or be muted completely.

 

[LeadingHeat]

With all that alleged machine-learning power of Ax SoCs, I wonder if Siri ever will be able to process language on-device, instead of requiring internet connection even for simple things like setting a timer, creating a reminder or playing a playlist.

I’m waiting patiently for this. I shouldn’t have to have an internet connection to set an alarm.

 

[The Game 161]

Ecosystem gets even better

 

[RoboCop001]

On real intercom you have to push a button to reply. This is no different.

for HomePod it is not recorded though. It is the records when you try to listen to the message on iPhone.

For me personally, it wouldn't feel the same. A button for me would feel faster than having to remember to first say "Reply" before my actual reply. But that's me. I feel like they could provide options at least. I'm sure it's not a big deal though and I'd probably get used to it. But to me it would feel odd to remove the emotion when saying "reply" and then putting the emotion back in when you say your actual message, lol.

 

[matrix07]

No it’s not. It sends the audio the same way you can send a voice clip in iMessage. That’s not real time. It’s quick, but not real-time.

HomePod to HomePod it’s real.

 

[matrix07]

Does the recipient have to accept and play the message? If so then it is different. Real time means the recipient hears the message at the same time that you speak it, which is what real intercoms do. I’m hoping that’s what this does, otherwise it’s just an audio iMessage conversation via Siri which already exists.

Yeah if you want to be too technical you can say it’s not real time since you have Siri “send the message for you”. You didn’t say “Hey Siri intercom..” and Siri on every HomePods in your house will stop all the music and be ready for intercoming, which will be ridiculous. You still get the message at that time, but not at the exact time the sender say, so you can response in (somewhat) real time. The most important thing is not it’s “exactly” real or not but is it like using intercom. For me, from the demo in the event, it’s definitely like. You said something. Everybody in the house heard it in your voice, not Siri voice, and can response. You can say it’s iMessage audio but that is no different than real time intercom with side benefit that you can also get alerts on your devices, so it’s actually better.

 

[sideshowuniqueuser]

"Hey Siri, tell everyone, dinner's ready."

Siri's reply: "Just a moment, I'm on it..."

After waiting for Siri for 30 seconds: Mom yells through the hall: "Dinner's ready!"

Then 5 mins after everyone has finished dinner...
Siri: "Dinner is ready!"

 

[DeepIn2U]

Very very neat.

Hopefully it works much better than the Walkie-Talkie feature, which fails to work most of the time.

you know a LOT of kids with TikTok’n parents are so going to HATE-this feature and be plugging out their pods when friends are over.

I cannot WAIT for the next round of TikTok videos with intercom video bombs leading to body face plant bombs (dance fails) lol.
So can’t wait 😄

 

[cmaier]

No it’s not. It sends the audio the same way you can send a voice clip in iMessage. That’s not real time. It’s quick, but not real-time.

It is not the same as a a voice clip in imessage. I just tried it. I said “hey siri tell everyone hello” into my iphone running beta 14.2, and as soon as i finished speaking, my voice came out of my HomePod running 14.2 in the other room.

 

[DeepIn2U]

My BF and I do as well. His office is in the house next door and it really comes in handy. The examples were all "tell everyone", but that makes me think there is a "tell someone" mode that will work with the U1 chip to only message the room/location/device needed to talk to that person.

I also think that is a precursor to two-way conversations (one-to-one is easier to implement).

oooh I see what you’re thinking. Speaker com safe sex late night huh? trust me don’t don’t it. Unless you want to be like that European TV commercial with that guy begin to have sex with his date on her door step with the intercom on and the older sister comes out to embarrass them all with a message from their dad!

Google tube the video you’ll crack up and see just how wrong your idea although VERY novel could go!

 

[insomniac86]

It is not the same as a a voice clip in imessage. I just tried it. I said “hey siri tell everyone hello” into my iphone running beta 14.2, and as soon as i finished speaking, my voice came out of my HomePod running 14.2 in the other room.

I didn't say it's exactly the same feature. As it autoplays the audio while iMessage won't.
But what you just said confirms that people are wrong and don't understand realtime.

You just said when you FINISHED speaking it played in other room. That is not real time just like how it works with iMessage voice clips. The data is sent complete not streamed realtime.

 

[yurc]

I assumed Mac compatibility available when Apple Silicon rolls.

 

[madeirabhoy]

It is not the same as a a voice clip in imessage. I just tried it. I said “hey siri tell everyone hello” into my iphone running beta 14.2, and as soon as i finished speaking, my voice came out of my HomePod running 14.2 in the other room.

isn't that confirming that its not real time?

 

[insomniac86]

isn't that confirming that its not real time?

Sure is lol. People really don’t seem to get it.

 

[cmaier]

I didn't say it's exactly the same feature. As it autoplays the audio while iMessage won't.
But what you just said confirms that people are wrong and don't understand realtime.

You just said when you FINISHED speaking it played in other room. That is not real time just like how it works with iMessage voice clips. The data is sent complete not streamed realtime.

True. Siri has to wait Until you finish the command, then it immediately send it. It’s not like “push to talk” on old cellphones.

 

[Shanghaichica]

This is even better than drop in or broadcast because it can be used across multiple devices away from home.

 

[Iconoclysm]

This is definitely a first world problem but I wish they would allow a different word besides Siri. The word Siri just sounds so unpleasant to the ears in my language that I am sure this is one of the reasons Siri has never really picked up here. Siiiiiiiiri just ugh lol it sounds so unfriendly and robotic. Alexa sounds better cuz it makes you feel like you are talking to a person (positive) and not a robot (negative association)

This is an understated request. There should be a command phrase that you can choose, not just a name.

 

[TLewis]

It IS real time conversation.

OK, I've only seen the event video, and haven't actually tried it, but this "real time" conversion seems awkward AF (and I really hope I'm wrong). Imagine this conversation:

Hey Siri, tell Sam, we need to go
Hey Siri, reply, just a moment.
Hey Siri, reply, come on, we need to go
Hey Siri, reply, I can't find my shoes
Hey Siri, reply, They're by the front door

Yeah, that's going to get old real quick. It'll also be interesting to see if Siri can handle those words said really quickly/loudly (e.g., when someone is annoyed/angry).

 

[matrix07]

OK, I've only seen the event video, and haven't actually tried it, but this "real time" conversion seems awkward AF (and I really hope I'm wrong). Imagine this conversation:

Yeah, that's going to get old real quick. It'll also be interesting to see if Siri can handle those words said really quickly/loudly (e.g., when someone is annoyed/angry).

Yeah using Hey Siri for long conversations will not cut it. Fortunately you will still be able to tap the top of HomePod and just say “Reply..” if you so choose, or I hope.
Apple should give us continuous Siri already.

 

[TLewis]

The thing is, a HomePod can already be a speakerphone. Apple just needs to make the software connect directly to one or more other HomePods and make them all speakerphones. This way, you just need to say "Hey Siri" once to make the connection, and people can just talk naturally. Once everyone stops talking for a couple of seconds or whatever, the HomePods can disconnect and go back to whatever they were doing. This would be a killer feature if you could make this hands-free without having to continually say "Hello Siri...".

(Apple's not stupid, and I'm sure they've thought of this. There's probably some privacy law somewhere that is preventing this.)

 

[Alan Wynn]

OK, I've only seen the event video, and haven't actually tried it, but this "real time" conversion seems awkward AF (and I really hope I'm wrong).

You are correct that would be annoying, but it does seem like the use case they were showing. What they were showing would be better called PA. It is not for conversations but broadcast announcements.

”Hey Siri, tell everyone dinner is ready.” does not require a reply. Once it gets coupled with U1 chips and good location information, it could enable open connections that were point to point and continuous.

”Hey Siri intercom with Sam ‘What do you want to do for dinner?’”
To which the reply could be:
“Hey Siri open intercom.” setting up an open channel between the two.

Just as my BF and I use Walkie-Talkie on our watches for very quick chats with low overhead:

”Where are you?” “What time are we leaving?” “Do you want to come with me to walk the dog?”

If these conversations get longer than one or two messages back and forth we will switch to FaceTime Audio.

 

[subjonas]

Yeah if you want to be too technical you can say it’s not real time since you have Siri “send the message for you”. You didn’t say “Hey Siri intercom..” and Siri on every HomePods in your house will stop all the music and be ready for intercoming, which will be ridiculous. You still get the message at that time, but not at the exact time the sender say, so you can response in (somewhat) real time. The most important thing is not it’s “exactly” real or not but is it like using intercom. For me, from the demo in the event, it’s definitely like. You said something. Everybody in the house heard it in your voice, not Siri voice, and can response. You can say it’s iMessage audio but that is no different than real time intercom with side benefit that you can also get alerts on your devices, so it’s actually better.

If someone is just making an announcement to the house, and maybe with one or two responses, the delay of recording and sending is fine. But what I’m wondering about is conversations. A real intercom doesn’t need to record; it’s real time so you can have free flowing conversation (granted you have to stand next to the intercom to press the button). But if Siri intercom has to record then send audio, then the same conversation would take twice as long, and in between speaking and listening you’d have a rather long silence which would impede on the flow of the conversation. Not to say it wouldn’t be useful, just not as useful as a real intercom, but but maybe there’s a technical limitation which is the reason for that. The most ideal would be to make it function like a phone conversation, with simultaneous two-way real time audio.