Paid Apple Intelligence features won't come until at least 2027

AppleInsider

In the forthcoming Apple Intelligence features, there was no mention of any fees or subscription charges. That could change as early as 2027.

In years to come, some advanced Apple Intelligence features could have a price tag attached.

Apple has not specified when all of the first features of Apple Intelligence coming in iOS 18, iPadOS 18 and macOS Sequoia will arrive. While the OS upgrades will be unveiled this spring, some of the AI features will be added to future updates spread across late 2024 and into early 2025.

AppleInsider has previously discussed that future Apple Intelligence features, particularly those for more advanced users, could come with a subscription or fee structure. That said, it is unlikely that users will see such features before 2027 at the earliest, according to Bloomberg's Power On newsletter.

The first set of features of Apple Intelligence include a combination of intelligent summarization abilities in programs like Mail and Messages, an improved Siri assistant, image generators, and more.

As with several of Apple's "add-on" services, such as Apple TV+ or iCloud+, more advanced features of Apple Intelligence are likely with a price tag. These features would likely be for specialized areas, or for large-scale services leveraging the Private Cloud Compute technology.

Subscription or paid features of Apple Intelligence are most likely be aimed mostly at corporate users, the education sector, or other business or specialist users. Some analysts have speculated that such a subscription could cost users as much as $20 per month.

Rumor Score: Likely

Read on AppleInsider

DAalseth

My only fear with this is that Apple might say Pages and Numbers are now enhanced with AI features and so make them subscription as well. MS has made a lot of money with Office365, why not do the same with iWork. 

Stabitha_Christie

DAalseth said:

My only fear with this is that Apple might say Pages and Numbers are now enhanced with AI features and so make them subscription as well. MS has made a lot of money with Office365, why not do the same with iWork. 

I doubt this will happen. Pages, Keynote, Numbers, iMovie, Photos and GarageBand are all part of Apple’s whole widget strategy. You can buy a Mac and be productive right out of the box. No need to buy anyone else. And if you need more advanced features there other apps you can buy. They might come up with AI features that you have to subscribe to get but charging for the actual app? Totally not going to happen. 

hypoluxa

I really hope Apple doesn't do any subscription based model for their AI features. They make enough $ charging a premium for their devices. Don't screw their user-base over like MS and become even more greedy. If they continue to make good products/services they'll make their money. They don't need to force users to a subscription service for a built in OS feature, IMO. We'll see though. I am thinking of upgrading to the iPhone 16 when it arrives (from my 12) to have the new AI features. I'd be PISSED if they were to force me to pay even more $ to use certain things that should be standard.

AniMill

I subscribe to Apple One. I don’t play any games or care about fitness tracking, but it’s still a good deal. IF they add AI to One, great. If they charge a tiny bit more, OK. If they creat a new tier, I won’t care. AI is a big nothing burger for me. 

slow n easy

I use ChatGPT because both Siri and Google are horrible at answering my questions and ChatGPT does a better job even though it’s not perfect. I have no idea about 4.0 because I’m not willing to pay for it and I’m not expecting to pay for Apple Intelligence once Apple decides to charge for it. I definitely will be looking forward to it though, if it’s for no additional cost. I have no idea if Apple intends to fix Predictive Typing with Apple Intelligence, but I wish they would. In my opinion, Predictive Typing does a horrible job. It changes my typing mistakes into nonsensical words and also actual words that make absolutely no sense within the context of the sentence.

eightzero

I see we now have entered the future where we predict the price of a product before anyone knows what it does or what it is for. 

williamlondon

eightzero said:

I see we now have entered the future where we predict the price of a product before anyone knows what it does or what it is for. 

And of course condemn one way or the other like the prediction is reality.

dutchlord

DAalseth said:

My only fear with this is that Apple might say Pages and Numbers are now enhanced with AI features and so make them subscription as well. MS has made a lot of money with Office365, why not do the same with iWork. 

Who uses iWork anyway? Totally unusable in a corporate environment. 

mpantone

This is all just pure speculation, there is no meat behind this rumor so it's really pointless to get all worked up about it.

For sure if Apple wants widespread adoption of Apple Intelligence they will need to offer some functionality free of charge (or offer a trial period). But the fact that they are going to launch Apple Intelligence without any fees would make it highly unlikely that they will erect a paywall around it (i.e. start charging for something they previously offered for free).

If Apple can only offer something comparable to what the competition offers for free, it will be hard for them to charge anything. The opportunity is if they compete with other paid services (like current ChatGPT+) or offer something differentiated that is head and shoulders above the competition.

And if Apple does create a paid level for Apple Intelligence, most likely they will offer some sort of free trial period. They have done this for pretty much all of their paid service offerings.

mattinoz

The obvious way for Apple to make additional revenue is allowing third party upselling from the basic functionality by allowing them to run custom and tuned models on Apples hardware in return for a cut of subscription.

You know basically the same model that has driven massive hardware sales and AppStore revenue for over a decade. 

blastdoor

mattinoz said:

The obvious way for Apple to make additional revenue is allowing third party upselling from the basic functionality by allowing them to run custom and tuned models on Apples hardware in return for a cut of subscription.

You know basically the same model that has driven massive hardware sales and AppStore revenue for over a decade. 

I think this is a good example of what it might look like. Another way to put it is that Apple would charges developers for using PCC in their apps. 

mattinoz

blastdoor said:

mattinoz said:

The obvious way for Apple to make additional revenue is allowing third party upselling from the basic functionality by allowing them to run custom and tuned models on Apples hardware in return for a cut of subscription.

You know basically the same model that has driven massive hardware sales and AppStore revenue for over a decade. 

I think this is a good example of what it might look like. Another way to put it is that Apple would charges developers for using PCC in their apps. 

Which could expand beyond AI uses to potentially any existing AS code that can handle the latency of being remote. 

Spanading_returns

I am going to wager that most of that is to do with the link in with ChatGPT as that is the cost of the subscription version. Hence, I can see this being an “add-in” for those that pay the subscription, so that ChatGPT can be accessed when the request is too much for AI. I think Microsoft has a similar model with their service. 

gatorguy

slow n easy said:

I use ChatGPT because both Siri and Google are horrible at answering my questions and ChatGPT does a better job even though it’s not perfect. I have no idea about 4.0 because I’m not willing to pay for it and I’m not expecting to pay for Apple Intelligence once Apple decides to charge for it. I definitely will be looking forward to it though, if it’s for no additional cost. I have no idea if Apple intends to fix Predictive Typing with Apple Intelligence, but I wish they would. In my opinion, Predictive Typing does a horrible job. It changes my typing mistakes into nonsensical words and also actual words that make absolutely no sense within the context of the sentence.

https://arstechnica.com/information-technology/2024/08/chatgpt-unexpectedly-began-speaking-in-a-users-cloned-voice-during-testing/
"Slow and easy" instead of all men to battle stations as ChatGPT seems to be doing.

Stabitha_Christie

dutchlord said:

DAalseth said:

My only fear with this is that Apple might say Pages and Numbers are now enhanced with AI features and so make them subscription as well. MS has made a lot of money with Office365, why not do the same with iWork. 

Who uses iWork anyway? Totally unusable in a corporate environment. 

What a painfully myopic comment. 

There are millions of consumers that buy Macs and iPads for personal use and don’t need Office.  There are millions of students that have Apple products that really don’t need office. 

Also, Apple uses Keynote for their keynote presentations. And Apple is ….. checks notes …. A corporation. So, there is at least one corporate environment that uses iWork and it doesn’t seem to hold them back in terms of success. 

jdw

ChatGPT4o, is the most brain dead stupid thing to come along in quite a while.  Sure, it does a few things surprisingly well, but it makes enormous mistakes of epic proportions every single time I use it. And when I attempt to correct it, it foolishly apologizes and then rephrases its stupid errors.  When I tell it repeatedly that it is in error, it repeatedly apologizes but never learns.

When I use ChatGPT via Bing (which I often do after my freeloader time limit for GPT4o runs out), it too makes mistakes. And when I get pissed off at it and give it a piece of my mind, silly Bing suggests I change the topic and a new session begins.  Totally hilarious that Bing is protecting ChatGPT!

Every single time I use ChatGPT4o in Chrome (which is the most reliable browser for it), at times when it gives me a clickable link, the link will never open when I click on it.  That forces me to tell it to give me links in plain text.  Why plain text?  So I can copy/paste it into a browser's address bar myself.  That's a work-around the "can't click the text link" bug, but quite nearly 100% of the time, when I finally am able to open the link, it yields a 404 or totally unrelated info.  It's maddening!

I am also not inclined to trust ChatGPT because it lies.  Because of it's blatant lies, I almost always ask it for links to its source info.  And that is how I know the links are bogus.  I sometimes believe what it tells me is true, but without a link to source info, I cannot trust it 100%.  I become especially doubtful when it gives me a link that leads to totally unrelated topics.

I mainly want to use ChatGPT to search the web faster and easier than I can Google something, asking it to check multiple sources.  Sadly, most of the time GPT4o misses info I know exists.  I often test ChatGPT based on what I know is out there on the web and is easily found by Googling.  Often times it misses that info when it does its own searches, and then comes back and tells me something wrong or incomplete.

I truly hate ChatGPT a lot of the time because it lies so often.  For example, I've recently been comparing OLED TVs and soundbars.  It tells me certain specifications that simply are not true.  When I point out its error, it apologizes and deletes the single error line, and then just repeats the rest it told me before.

Why do I even torture myself by continuing to use the stupid thing?  Because it sometimes does a decent job in very specific situations like summarizing text or rephrasing.  I prefer that use case because I am good enough in the English language to know if what it tells me is good or not.  But in other cases where I am looking for facts I don't know to be true, it often lies; and even when it doesn't lie, I try to get links to source info, but those links never work.

So it's a real love-hate relationship, but more hate than love.

It is totally and utterly laughable that governments and people around the world are afraid of AI.  Yeah right!  Maybe 100, 200 or perhaps 300 years from now it might be worthy of such fears, but at the moment it's not that far ahead of brain dead Siri.  It's no different than the fake promise of "self driving cars." That's not going to happen in my lifetime.  Truly autonomous driving means no human intervention is EVER needed, and you can drive in any situation, like mountain driving in the snow with sunlight reflections hitting the snow and blinding you occasionally.  Or driving on dark roads that aren't marked with paint.  Or driving off-road in the dirt.  Nope.  No matter what these car companies say, what they have now is little more than a joke.  It gives you a great first impression, but using the tech for a while shows how ridiculous it is.  

The PROMISE of AI is great, but it's far, far in the future before we can sit back and experience true "intelligence" that's "artificial."  What we have now in the world of AI is barely useable.  It's more entertainment than anything else.

I write all this to say that if Apple CHARGES MONEY for such digital stupidity, I certainly will not be lining up for a subscription, They have a LONG way to go before its worthy of dedicated fees and special charges.  We're very much in the R&D stage right now.  I say this in the hope of something far superior to what we have now, while at the same time, I am a realist too.  It's not great now, and it probably still won't be that great 10 to 20 years hence.  We need a real technological breakthrough to make a huge leap in usability.

blastdoor

mattinoz said:

blastdoor said:

mattinoz said:

The obvious way for Apple to make additional revenue is allowing third party upselling from the basic functionality by allowing them to run custom and tuned models on Apples hardware in return for a cut of subscription.

You know basically the same model that has driven massive hardware sales and AppStore revenue for over a decade. 

I think this is a good example of what it might look like. Another way to put it is that Apple would charges developers for using PCC in their apps. 

Which could expand beyond AI uses to potentially any existing AS code that can handle the latency of being remote. 

Indeed — it could become Apple’s analog to AWS.

blastdoor

jdw said:

ChatGPT4o, is the most brain dead stupid thing to come along in quite a while.  Sure, it does a few things surprisingly well, but it makes enormous mistakes of epic proportions every single time I use it. And when I attempt to correct it, it foolishly apologizes and then rephrases its stupid errors.  When I tell it repeatedly that it is in error, it repeatedly apologizes but never learns.

When I use ChatGPT via Bing (which I often do after my freeloader time limit for GPT4o runs out), it too makes mistakes. And when I get pissed off at it and give it a piece of my mind, silly Bing suggests I change the topic and a new session begins.  Totally hilarious that Bing is protecting ChatGPT!

Every single time I use ChatGPT4o in Chrome (which is the most reliable browser for it), at times when it gives me a clickable link, the link will never open when I click on it.  That forces me to tell it to give me links in plain text.  Why plain text?  So I can copy/paste it into a browser's address bar myself.  That's a work-around the "can't click the text link" bug, but quite nearly 100% of the time, when I finally am able to open the link, it yields a 404 or totally unrelated info.  It's maddening!

I am also not inclined to trust ChatGPT because it lies.  Because of it's blatant lies, I almost always ask it for links to its source info.  And that is how I know the links are bogus.  I sometimes believe what it tells me is true, but without a link to source info, I cannot trust it 100%.  I become especially doubtful when it gives me a link that leads to totally unrelated topics.

I mainly want to use ChatGPT to search the web faster and easier than I can Google something, asking it to check multiple sources.  Sadly, most of the time GPT4o misses info I know exists.  I often test ChatGPT based on what I know is out there on the web and is easily found by Googling.  Often times it misses that info when it does its own searches, and then comes back and tells me something wrong or incomplete.

I truly hate ChatGPT a lot of the time because it lies so often.  For example, I've recently been comparing OLED TVs and soundbars.  It tells me certain specifications that simply are not true.  When I point out its error, it apologizes and deletes the single error line, and then just repeats the rest it told me before.

Why do I even torture myself by continuing to use the stupid thing?  Because it sometimes does a decent job in very specific situations like summarizing text or rephrasing.  I prefer that use case because I am good enough in the English language to know if what it tells me is good or not.  But in other cases where I am looking for facts I don't know to be true, it often lies; and even when it doesn't lie, I try to get links to source info, but those links never work.

So it's a real love-hate relationship, but more hate than love.

It is totally and utterly laughable that governments and people around the world are afraid of AI.  Yeah right!  Maybe 100, 200 or perhaps 300 years from now it might be worthy of such fears, but at the moment it's not that far ahead of brain dead Siri.  It's no different than the fake promise of "self driving cars." That's not going to happen in my lifetime.  Truly autonomous driving means no human intervention is EVER needed, and you can drive in any situation, like mountain driving in the snow with sunlight reflections hitting the snow and blinding you occasionally.  Or driving on dark roads that aren't marked with paint.  Or driving off-road in the dirt.  Nope.  No matter what these car companies say, what they have now is little more than a joke.  It gives you a great first impression, but using the tech for a while shows how ridiculous it is.  

The PROMISE of AI is great, but it's far, far in the future before we can sit back and experience true "intelligence" that's "artificial."  What we have now in the world of AI is barely useable.  It's more entertainment than anything else.

I write all this to say that if Apple CHARGES MONEY for such digital stupidity, I certainly will not be lining up for a subscription, They have a LONG way to go before its worthy of dedicated fees and special charges.  We're very much in the R&D stage right now.  I say this in the hope of something far superior to what we have now, while at the same time, I am a realist too.  It's not great now, and it probably still won't be that great 10 to 20 years hence.  We need a real technological breakthrough to make a huge leap in usability.

 TL;DR

muthuk_vanalingam

blastdoor said:

jdw said:

ChatGPT4o, is the most brain dead stupid thing to come along in quite a while.  Sure, it does a few things surprisingly well, but it makes enormous mistakes of epic proportions every single time I use it. And when I attempt to correct it, it foolishly apologizes and then rephrases its stupid errors.  When I tell it repeatedly that it is in error, it repeatedly apologizes but never learns.

When I use ChatGPT via Bing (which I often do after my freeloader time limit for GPT4o runs out), it too makes mistakes. And when I get pissed off at it and give it a piece of my mind, silly Bing suggests I change the topic and a new session begins.  Totally hilarious that Bing is protecting ChatGPT!

Every single time I use ChatGPT4o in Chrome (which is the most reliable browser for it), at times when it gives me a clickable link, the link will never open when I click on it.  That forces me to tell it to give me links in plain text.  Why plain text?  So I can copy/paste it into a browser's address bar myself.  That's a work-around the "can't click the text link" bug, but quite nearly 100% of the time, when I finally am able to open the link, it yields a 404 or totally unrelated info.  It's maddening!

I am also not inclined to trust ChatGPT because it lies.  Because of it's blatant lies, I almost always ask it for links to its source info.  And that is how I know the links are bogus.  I sometimes believe what it tells me is true, but without a link to source info, I cannot trust it 100%.  I become especially doubtful when it gives me a link that leads to totally unrelated topics.

I mainly want to use ChatGPT to search the web faster and easier than I can Google something, asking it to check multiple sources.  Sadly, most of the time GPT4o misses info I know exists.  I often test ChatGPT based on what I know is out there on the web and is easily found by Googling.  Often times it misses that info when it does its own searches, and then comes back and tells me something wrong or incomplete.

I truly hate ChatGPT a lot of the time because it lies so often.  For example, I've recently been comparing OLED TVs and soundbars.  It tells me certain specifications that simply are not true.  When I point out its error, it apologizes and deletes the single error line, and then just repeats the rest it told me before.

Why do I even torture myself by continuing to use the stupid thing?  Because it sometimes does a decent job in very specific situations like summarizing text or rephrasing.  I prefer that use case because I am good enough in the English language to know if what it tells me is good or not.  But in other cases where I am looking for facts I don't know to be true, it often lies; and even when it doesn't lie, I try to get links to source info, but those links never work.

So it's a real love-hate relationship, but more hate than love.

It is totally and utterly laughable that governments and people around the world are afraid of AI.  Yeah right!  Maybe 100, 200 or perhaps 300 years from now it might be worthy of such fears, but at the moment it's not that far ahead of brain dead Siri.  It's no different than the fake promise of "self driving cars." That's not going to happen in my lifetime.  Truly autonomous driving means no human intervention is EVER needed, and you can drive in any situation, like mountain driving in the snow with sunlight reflections hitting the snow and blinding you occasionally.  Or driving on dark roads that aren't marked with paint.  Or driving off-road in the dirt.  Nope.  No matter what these car companies say, what they have now is little more than a joke.  It gives you a great first impression, but using the tech for a while shows how ridiculous it is.  

The PROMISE of AI is great, but it's far, far in the future before we can sit back and experience true "intelligence" that's "artificial."  What we have now in the world of AI is barely useable.  It's more entertainment than anything else.

I write all this to say that if Apple CHARGES MONEY for such digital stupidity, I certainly will not be lining up for a subscription, They have a LONG way to go before its worthy of dedicated fees and special charges.  We're very much in the R&D stage right now.  I say this in the hope of something far superior to what we have now, while at the same time, I am a realist too.  It's not great now, and it probably still won't be that great 10 to 20 years hence.  We need a real technological breakthrough to make a huge leap in usability.

 TL;DR

Well, I did read it and found it providing useful information.

dewme

williamlondon said:

eightzero said:

I see we now have entered the future where we predict the price of a product before anyone knows what it does or what it is for. 

And of course condemn one way or the other like the prediction is reality.

Spot on from both of you. I’m still at a place where I don’t truly understand what Apple Intelligence really is on a personal level. Why speculate about subtle nuances like how it may possibly be monetized by Apple until we truly understand what it is or what value it actually brings to our personal lives. 

It may turn out to be wildly life changing, a total flop, or something in-between. I won’t know until I’ve tried it. There have been plenty of “amazing new features” launched by Apple over the years that turned out to be totally meh. 

I think Apple has pretty much refined all of the basic features of all of their premier products to the point where they mostly “just work.” They still have work to do in some areas. Some things have turned out to be dangerously too good, e.g., as evidenced by a large proportion of smartphone users being psychologically addicted to their smartphones at the expense of properly dealing with the physical things and people in their non-virtual lives. At this point the added value of new features is very incremental despite all of the hype behind them. 

Will Apple Intelligence fundamentally change the way we use Apple products or will it be just another layer of features on top of an already tall stack of features? I have no idea. So far I’ve been impressed with ChatGPT being a “better search and query” tool that may occasionally return bogus answers. 

Would I pay to use ChatGPT in its current form? Probably not only because there are free tools that can provide most of the things I use ChatGPT for. Convenient? Yes. Essential? Not yet. I still have a reasonably functioning brain and can generate most of the things I need to generate using the grey blob the sits between my ears. 

This article may be wildly speculative, but it does illustrate a real concern about how AI in general will pay for itself. Will it be used to raise the quest for answers and knowledge for everyone or will it follow the prevailing capitalist tendencies in countries like the US to only provide the very best benefits to those who can afford to pay, in a similar manner to US healthcare?

It will be very interesting to see which direction Apple goes with Apple Intelligence. Raise all boats or pay to play?