Here's a First Look at Apple Music's Voice Plan Launching With iOS 15.2

[RyanSpecter]

After some serious mental gymnastics, I think that fundamentally the plans are fine, and they're likely targeted at users of Spotify and other music apps that haven't themselves implemented the Siri Support on the HomePod. The major problem with it is the poor messaging around what exactly it is. If Apple said "This is for you folks that still don't have Spotify on your HomePods", people would say. "Woah, if this was $2 a month, that would be pretty neat!"

I imagine some PR guy at Apple said, "we can't outright slate Spotify for not implementing their app on our smart speaker, so we have to be careful in what we call this thing"

 

[ipedro]

Incredibly dumb looking interface. However, I understand that there’s a teeny-tiny market of users who only access Apple Music through smart speakers and that this is enticing for small business owners to make the Pandora -> voice plan shift. Still, who thought this was Apple Music’s top priority?

It’s not as “teeny-tiny” as you think. I’m an avid Apple fan, use Apple products every day, both professionally and at home but I rarely ever open Apple Music on my iPhone, iPad or Mac to type in a song or curate playlists.

That’s just not how I listen to music.

I walk into a room and ask a HomePod to play music in that room or everywhere or I lift my wrist when outside and ask the same to my Watch. I’m not asking for specific songs, I just want to listen to music. Apple Music learned about my tastes through skips, and ”Hey Siri, I love this song”. I remember it taking maybe a week or two before almost all the music played were songs that I either already liked or new music that fit my tastes. And not only are the selections great, Apple Music plays appropriate music for the time of day.

While I won‘t subscribe to this lower tier since Apple Music is included in my Apple One plan, there are clearly a sizeable number of holdouts who still don’t subscribe to a music service, and that’s who this is targeting. So many people on this forum have incredibly narrow mindsets and can’t imagine anybody being different than themselves.

 

[ipedro]

…Then there are those who take the narrow mindset to an entirely different level, like this user who has gone and added angry emojis to every post that explains how there are people who might like this more basic service. ?

 

[mika-74]

Biggest problem of Siri is the recognizing of titles in other languages. It drives me crazy. If you use it in italian it is impossible to ask a german song. You must try to spell it in a stupid way and it mostly not work. Sometimes it works with english songs. If I set Siri’s language to english, it will not understand italian titles or singers. It simply won’t work!

 

[radioflyer]

This service ONLY works with Siri. You can’t browse music in the music app, or create playlists, etc. You must use Siri for EVERYTHING.

With that being said, instead of paying $9.99/month you’re only paying $4.99/month. So for the inconvenience of only being able to use Siri to play music, you save $5/month.

This might be worth it to some, I suppose, but I can’t imagine a lot of people signing up for this plan.

Thank you for clarifying ??

 

[Thomas Davie]

Me: Siri, play warrior on the edge of time by Hawkwind
It: Gives me Hawkwind by Hawkwind on…..

Me: Siri , play warrior on the edge of time by Hawkwind
It: Gives me Hawkwind: Choose your Masques by Hawkwind

Me: Swears
It: Gives me Poopoo peepee by Baby Kada?

Really puzzled at the last one. But I sent it to my niece’s kids and now they sing it all the time.

I am not against this new tier of Apple Music and can certainly see a segment of the market liking this….until they come face to face with the inaccuracies inherent in a voice controlled only music service. If I’m in the mood for truly random music I’d jump all over this.

Tom

 

[KindJamz]

…Then there are those who take the narrow mindset to an entirely different level, like this user who has gone and added angry emojis to every post that explains how there are people who might like this more basic service. ?

View attachment 1926792

But your stalking his history. Kinda weird.

 

[Denzo]

I’d rather pull my testicles through a straw than be bound to using Siri on a daily basis.

If it was Alexa or Google, I’d consider it, but Siri would be a nightmare for this function.

 

[Thomas Davie]

I’d rather pull my testicles through a straw than be bound to using Siri on a daily basis.

If it was Alexa or Google, I’d consider it, but Siri would be a nightmare for this function.

A lot of the mistakes I encounter are b/c I don’t setup playlists; even so it’s got to be something Alexa can understand.

Have never managed to get Alexa to correctly play the album Doremi Faso Latido (again by Hawkwind).

Siri does manage to get simple playlist names almost 100%. Wierd thing is the (Google assistance) is very accurate. So far it’s the only voice assistant that manages to understand ’play the album Senjutsu by Iron Maiden’

Tom

 

[jtdx99]

Shouldn't this be free? I'm a big fan of Apple but Siri honestly sucks, especially being a multilingual speaker.

 

[mctape]

Me: Siri, play warrior on the edge of time by Hawkwind
It: Gives me Hawkwind by Hawkwind on…..

Me: Siri , play warrior on the edge of time by Hawkwind
It: Gives me Hawkwind: Choose your Masques by Hawkwind

Me: Swears
It: Gives me Poopoo peepee by Baby Kada?

I think it's funny that people crap on Siri for doing this when this is exactly what it sounds like when my roommate is trying to play any song through his Alexa device.

 

[kc9hzn]

I think it's funny that people crap on Siri for doing this when this is exactly what it sounds like when my roommate is trying to play any song through his Alexa device.

Right, music (especially individual track) selection is basically an impossible problem for AI. It’s all contextless free text (conventional grammar rules for conversation fly out the window, for instance), there’s the problem of cover versions, similarly titled songs/artists, subtly different cuts of the same song on different albums, and then there’s the issue of long tail optimization. I’m sure the base of the tail (think, current top 40 radio, a la the “that’s so sad, Alexa, play Despacito” meme) is super optimized (as, presumably, those are the most common requests and the most frequently played songs), while the long tail (the songs that only rarely get voice requests or the songs that literally never get voice requests) is probably far less well tuned. And this is true for any AI assistant. Another “fun” category of requests is trying to request playing a specific episode of a given podcast, for much the same reasons.