Very Apple-ish approach to AI catch up: wrap an external tool in a privacy architecture, embed into the OS and productize the orchestration layer.
It will be interesting to see if the Private Cloud Compute + on-device routing can make third-party model capabilities feel like a first-party system without leaking user context to the model provider.
If Apple handles the Google-Apple boundary right, this will be an elegant move on their part, otherwise it will feel like Apple Intelligence with a just a privacy-polished frontend for Gemini.
Wouldn't it be more accurate to call Apple's architecture data protection rather than privacy? As an European citizen in a post Snowden world I would be surprised if any of my data on Apple services was actually kept private from the US government, and Apple certainly wants to own a lot of data/metadata about you. Gotta have Siri listening for carplay and so on. I would aboslutely trust Apple not to sell my data as a commodity though.
> If Apple handles the Google-Apple boundary right, this will be an elegant move on their part, otherwise it will feel like Apple Intelligence with a just a privacy-polished frontend for Gemini.
I'd say this is spot on. At least if what Microsoft is doing with Copilot Cowork is anything to go by. Cowork is not a privacy-polished as much as it's an Enterprise compliant polish to make Opus 4.8 run "safely" in your enterprise organisation. So far Microsoft is winning the AI war in non-tech enterprise with this, especially here in the EU. If Apple manages to do this for the private market that will be great for them.
I'm not personally sold on what an AI should do on my phone though. I use a lot of AI professionally, but I haven't even turned on Bixby or whatever the Samsung AI is called.
From an EU perspective, Microsoft is doing data protection, Apple is doing data privacy.
Microsoft's approach to data is basically "we promise nobody else but you and your government can access it, we can but we pinky swear we won't." This promise is mostly enforced at the legal layer and through legal consequences, not technical safeguards. If they think they can get away with it (or are forced to get away with it by the US government), there's nothing stopping them from using your data in whatever way they want.
When they can, Apple designs their systems so that they physically don't even have the capability to use your data, even if it's processed on their own servers. They're not privacy maximalists like Signal is, they care more about user experience, but they do aim for the highest level of privacy you can get while still having a good experience, and when they do need to make sacrifices, they typically let you opt into the privacy features if you really want to.
I'm far more inclined to believe that Microsoft is secretly (or not so secretly) collaborating with the US government than that Apple is.
> Wouldn't it be more accurate to call Apple's architecture data protection rather than privacy? As an European citizen in a post Snowden world I would be surprised if any of my data on Apple services was actually kept private from the US government, and Apple certainly wants to own a lot of data/metadata about you.
Your conception doesn’t seem to match PCC at all. The whole point of it is that nobody can access the data, not even the people running the servers.
> I use a lot of AI professionally, but I haven't even turned on Bixby or whatever the Samsung AI is called.
I know this was just a small aside, but man do I hate Bixby and other phone AIs. They are so frustratingly difficult to turn off, and turning them on accidentally is as simply as holding the wrong button for a few seconds, such as when your phone is in your pocket. Very frustrating design.
> I would be surprised if any of my data on Apple services was actually kept private from the US government
Outside of law enforcement having a warrant, Apple's efforts against CSAM, or their Chinese data centers, I've not heard of Apple doing any of what you assume in a post-Snowden world. iMessage is supposed to be end to end encrypted, and there was a few years ago that whole scandal where Apple wouldn't unlock a literal terrorists cell phone for the FBI.
The FBI had to reach out to... a third party to unlock the phone (I forget the name of the firm that did it - Cellebrite maybe?) for them, what's funny is they spent a lot of money on it, when the rest of the world pointed out that the very specific iOS version in question had known vulnerabilities they could have found online for free (or cheaper?).
Is there a meaningful Google-Apple boundary in operation?
They are buying the right to distill their own Gemini models and run them in their data centres (or at least data centres they control); unless I am missing something, this isn't going to be infrastructure that Google has operational control over.
If Apple is running the inference from Apple iPhones and Apple data centers then Apple has operational control. Google’s influence ends the moment they hand the weights over to Apple.
As someone who doesn't use Android, they showed a lot of integration into the apps, which I think is where the real magic happens, and it's not something I can do with any 3rd party chatbots today (that I'm aware of). I also don't know that I would trust the other 3rd parties with the access required to pull it off.
I'm interested in how it feels to use: whether there is any context leaking, as you mentioned, if it introduces latency, and whether there are any pricing implications? I know they weighed a variety of factors, including the smaller models, but cost had to be a big concern, too... I feel like Google is the only provider giving away so much AI inference for free.
This will further blur the picture about when and how consumers / employees are supposed to pay for AI services. For example, they showed consumer rather than coding tasks, but could you select five files and ask Siri to write a Python script or a small app? Will enterprises just disable Siri AI functionality, or will they be able to route it through their own AI auditing and providers?
Consider that they described the development process as taking Siri down to bare metal and rebuilding on a new architecture. I don’t think our previous Siri experience will be particularly relevant.
I absolutely adore the historical revisionism that apple cares about privacy.
Run your router through a linux laptop as a proxy so you can capture traffic, connect any apple device to your router, and see the vasts amount of data your device sends to apple.
Apple DGAF about privacy, they want your data as much as anyone else, their only thing is that they should be the only ones to get it and then other people have to pay them for it, rather than your device sending the data to the 3d party directly.
And if you think your data is secure, reminder that The Fappening was all done targeting apple devices.
I would love to learn more about what's actually powering Apple Intelligence now. Are they using flagship Gemini models behind their own prompts? Fine-tuning? Pre-training their own models based on Gemini?
Is there a meaningful distinction between the Gemini-powered models and Apple Foundation Models? Does that distinction vary for on-device vs hosted models? Are some models running on Apple's Private Cloud Compute and others running on Google iron?
Edit: they elaborated significantly in a "keynote tech-talk": [0]
According to Apple, there are five models:
On-Device
- AFM Core: Dense architecture; the standard next-gen on-device model
- AFM Core Advanced: Sparse architecture, natively multimodal; enables features like image understanding and expressive voices
Private Cloud Compute
- AFM Cloud: Workhorse server model optimized for latency and cost
- AFM Cloud Image: Image generation and editing
- AFM Cloud Pro: Most capable model, Gemini frontier-level quality, for complex reasoning and agentic tasks; runs on NVIDIA GPUs in Google's cloud under Apple's PCC privacy guarantees
Everything excluding Cloud Pro are custom models running on Apple Silicon, "refined" using Google Gemini. About Cloud Pro, they say "this is our most capable model with quality similar to Gemini frontier models." So I might read between the lines and say this is a wrapped Gemini.
> Are they using flagship Gemini models behind their own prompts? Fine-tuning? Pre-training their own models based on Gemini?
Easiest way to tell, how much dancing around did we listen to and how many diagrams did we have to look at? If they had their own tech we wouldn't be looking at diagrams we'd just be getting told Siri AI, it's private, it's powerful, here's what it can do. Instead we had 10 minutes talking around the tech and this diagram [1] which is a signal that it's a bunch of other peoples stuff cobbled and wrapped together.
Local is probably similar to Gemma e4b you can get right now on Google Edge Gallery (the ios and Android app). Guessing that the more powerful version that will only work on the 12gb ram devices will be something unreleased that is similar but a bit larger
Google also awhile back announced being able to run full Gemini by leasing / renting hardware in your own datacenters so companies can train or access data without needing to send things to their datacenters. Nvidia based. Guessing Private Compute might just be Apple leasing a ton of those?
Am I reading this correctly? Their chosen cloud providers run the PCC stack on their hardware, so the compute provider is responsible for ensuring the privacy guarantees? I assume that would add to the potential security surface area.
The public version of Gemini is ridiculous. At least half their search "answers" are just wrong. If you then start a follow up chat the answers change but usually still half wrong.
Search would be better without the added AI hallucinations above it. If I want an AI answer I'll go and ask Claude, the quality difference is huge.
It's strange to me that Apple would choose to disadvantage themselves by selecting Google as their provider as opposed to, say, Anthropic or even OpenAI. Doesn't this mean they'll struggle more to differentiate themselves from the assistant on Android phones? Thinking more cynically, couldn't Google, if they wanted, feed Apple an inferior version of Gemini, ensuring they stay ahead?
As the consumer, this just sucks because it means no matter which phone platform you choose, you're getting the same thing underneath, and there's no way to avoid it (besides not using an assistant entirely, which I recognize a lot of people do, myself included).
I'd wager that for 99.9% of "Apple Intelligence" tasks, Google's models perform just as well as other frontier labs. Google also has done more work on getting LLMs running on edge devices compared to anthropic and openAI.
The source also says
> The new architecture centers on Apple Foundation Models co-developed with Google, which Apple says are adapted to run both on-device and on servers through its existing Private Cloud Compute infrastructure
Which could mean Google and Apple have trained some custom models, probably the on-device ones, specifically tailored towards Apple's hardware.
That's not the point of OP. The point of OP is that Google is a direct competitor of Apple for phones / OS, so them giving the "key of the house" to Google is risky
I see it differently… Apple has chosen to treat the model as a commodity. By making Gemini an implementation detail, they leave the door open to swap it out for Anthropic or OpenAI without end-users even knowing or caring. So I think they're creating leverage in any future negotiation.
And at the same time they claim it can't be rolled out in Europe because the DMA would force them to allow selection of other AI providers? Which would not even be true if this was an Apple product built on the models of Google, just like the DMA does not force them to pick a different datacenter of office cleaning provider.
> I see it differently… Apple has chosen to treat the model as a commodity
It didn't work out well for Yahoo.com. It turned out that Google Search was the value and yahoo.com just skin around it. It might be the same for Apple. Gemini is the valuable part, what particular device you use it on matters less.
In time Apple will swap it out for a in house model like many other things they’re swapped out in the last 25 years, Apple appears to be a company that doesn’t waste money and seem to execute long range projects if necessary I don’t think the Google models will be there for long. I think they will be swapped out when the M series GPUs get to the performance level they want.
Bingo. They see this as the future commodity it will be. Customers will choose AI providers much way they choose a car: taste, price, a few other factors.
And to your point, Google has a massive balance sheet, produces their own AI chips, and is not going anywhere anytime soon.
Google was the right choice stability and it only cost Apple $1 billion per year that’s pretty much of a no-brainer, and with Apple’s history, they probably will use Gemini for as long as they need it and then use their own model in time.
> It's strange to me that Apple would choose to disadvantage themselves by selecting Google as their provider as opposed to, say, Anthropic or even OpenAI. Doesn't this mean they'll struggle more to differentiate themselves from the assistant on Android phones? Thinking more cynically, couldn't Google, if they wanted, feed Apple an inferior version of Gemini, ensuring they stay ahead?
Is it really all that difference from Apple defaulting to Google's search engine?
> Is it really all that difference from Apple defaulting to Google's search engine?
Not really, because the business model isn't there (at least not in this iteration.
1. The models are Apple models, co-developed with Google. They are not white-label Gemini.
2. There's not currently a Google failover or UX
3. Because of that, there's no user monetization to share.
Apple does have a ChatGPT integration, with failover UX, and with a suspected revenue share deal. However, one could see this deal in a precarious situation, since at the time it started it was expected Apple would not focus much at all on a model capable with world knowledge.
Depends on which way the money is flowing. Google pays Apple for default search engine. Is Google paying Apple for using Gemini? That feels like a much heavier investment if they are
Yes, Google can do that just like Intel, Samsung, Nvidia or Qualcomm yes Google can drag their feet, we know in the end it will all lead to tears and then they will separate.
At the time Apple made this decision there wasn’t as strong of a difference in model quality between Google, anthropic, and OpenAI as there is now. Also Apple definitely burnt some bridges with OpenAI on the agreement they made together a year earlier.
OpenAI and Anthropic don't make small models. Google happens to already have a billion devices that would benefit from small models, so they made one.
Google basically gets 1 billion per year for free*.
Gemini is better than either at multi modal, google also has their tensor processor stuff with ridiculously high T/s output they need for acceptable UX
Only Google was willing to do what Apple wanted for a reasonable price.
They licensed Gemini and Google infrastructure not just for use, but to accelerate the creation of the three independent Apple Foundation Models announced today:
- AFM Core
- AFM Core Advanced
- AFM Cloud
Google also worked to be able to host AFM Cloud on their infrastructure per Apple's private cloud compute architecture, including some form of independent third party review/audit.
I suspect the only two organizations with both the model and the infrastructure needed for Apple were Google and xAI - and I'm not sure Apple would touch Grok with a ten foot pole, even if xAI were willing to let it be used for training.
Maybe. Search ads likely make Google more money than they pay apple. For AI, Google currently loses money. If they eventually make money via ads, then sure. Else, apple will have to pay them
Search has ads but Siri doesn't. And when Apple puts ads in Siri they won't be Google ads. I don't think Google benefits from this deal enough to be worth paying.
Google doesn't care they do not control iOS. Google cares to have their products everywhere. For many years Google apps were better on iOS than on Android, because on Android they just preinstall them while on iOS they need to be installed explicitly.
Android is not a Google product, it is just a tool for Google to collect data. If they manage to collect data via Apple Intelligence they are going to do it. Regardless of what Apple marketing says.
It may be a clever move. By using the same models as android (contractually?), they can compete on the user experience which they typically handle better than android phone providers.
Apple originally partnered with OpenAI. We won’t know all the details for some time, but given OpenAI’s penchant for drama (they started leaking that they might sue Apple [1]), it seems fair to sideline them as a long-term partner.
It's not clear that there's any sort of durable advantage to the provider here -- in fact, oAI started as the partner with apple a couple of years ago, and is reportedly unhappy with the outcomes.
The hard part is not distilling a frontier model down into a specific use case when you have hundreds of millions of users, the hard part is (apparently) re-architecting your mobile OS to work with such a model rather than fight it. Those architectural benefits accrue to apple, as will future datasets and expertise, and the benefit of having some distillation working 24/7 on prem.
Anyway, where I think you're going to be grumpy in two years is that switching the underlying model is going to require a jailbreak, and that you wish they'd made the os much more deeply open for agentic interaction, not that it's gemini - it's just not the valuable part of the story for Apple or for users.
The differentiation is the integration, not the model itself which is mostly fungible. And afaik Apple is running these models on their own compute, so I don't think google can pull a bait and switch.
> It's strange to me that Apple would choose to disadvantage themselves
How exactly are they disadvantaging themselves? Perhaps expand on that opinion a bit, and the data/assumptions you're making in forming it.
My naive assumption is that they're going to do what everyone is doing: make tooling that lets you swap in any model.
I don't think it's fair to assume incompetence, on their part. I think it's much safer to assume they're doing what's best for them, and it's very clear to them what's best (money, support, etc).
> My naive assumption is that they're going to do what everyone is doing: make tooling that lets you swap in any model.
They are a ways away from that for Siri, as they can't guarantee third party tooling meets their security and privacy requirements. Meeting those security and privacy requirements also makes it harder for a third party to monetize their investment or ongoing use of infrastructure.
But I suspect you will see integration in other areas, such as image generation.
> It's strange to me that Apple would choose to disadvantage themselves by selecting Google as their provider as opposed to, say, Anthropic or even OpenAI.
The models are quickly converging to similar capabilities with particular ones being better at a particular task until the next release cycle.
> Thinking more cynically, couldn't Google, if they wanted, feed Apple an inferior version of Gemini, ensuring they stay ahead?
Yeah, that's not really how things work at this scale.
Differentiating themselves can only be a bad thing for Apple. Google is way ahead of Apple in AI and likely always will be.
Both Apple and Google end up advantaged by this relationship: Apple gets the same technology as Android, meaning there is no competitive front opening up. Google gets eyeballs on almost the entire smartphone market.
Apple can continue to differentiate the iPhone from Android in all the ways that they were doing before.
The agent harness matters just as much as the AI model. Using Hermes or OpenClaw feels like night vs day when using OpenAI’s apps even when using the same exact model.
You can even see difference in agent harnesses using the same model in the same company if you compare Gemini CLI with AntiGrav. They are different experiences.
I’m pretty sure Apple’s agent harness will be drastically different from Google’s even with the same model
Anthropic or OpenAI have no foothold into the mobile market. Google has integrated a shitton of AI functionality into their latest Pixel phones. That’s what would scare me if I was Apple and worried that if AI prevails this could steal some market share from me. The other two are irrelevant in this context.
I'm speculating, but it's likely that Google is the only provider who is willing to adhere to the inference compute requirements that Apple sets out for their foundation models. They are, after all, the only provider that will let you host their FMs in your data center.
What part of using Gemini do you think is a disadvantage?
Also important to remember how immature OpenAI and Anthropic are as companies. It would be a huge technical, legal, and reputational risk to commit to using them.
Didn't they famously have a search deal with Google (that they were also ultimately fined for - in the EU at least)? So there's definitely precedent with Google as a "partner".
A functioning FTC would not allow this. Insane that there will be no competition for integrated smartphone AI because the existing duopoly at the OS level has agreed to team up.
>couldn't Google, if they wanted, feed Apple an inferior version of Gemini, ensuring they stay ahead?
Why are Apple people like this lmao. Yeah ofc they could, but they won't because 2 businesses as large as them have a deal it's usually honored.
Couldn't anybody in Apple's supply chain sabotage them? Contrary to popular belief, I don't think Google is in direct competition with Apple at all...Google don't make a mobile operating system. And they certainly don't make much hardware (especially that people should buy).
If I can't even trust the results given by ChatGPT and Claude at their highest level of reasoning in my daily life and work, would I be willing to use Siri AI to handle the important scenarios depicted in the livestream?
The idea that someone would go on a multi hour car trip based on an AI-generated route without manually checking where they're actually going is quite funny to me. Some older HN members will remember when the first stories a la 'the GPS told me to drive down that ravine' started coming out. I'm looking forward to "my AI assistant routed me to Kangiqiniq, Canada and border security detained me when I couldn't produce a passport".
I sometimes think about the time I was on a road trip with a friend, and for what was generally a 3 hour trip to the mountain, was 5 hours for my friend driving.
Midway through the trip I was suspicious of the duration, traffic was fine. He was adamant he knew where he was going.
I pulled up google maps, and sure enough. 3 hours.
Turns out, his mapping app wasn’t aware of an offramp, so instead it wanted us to drive an extra hour, then do a u-turn and drive back to take an offramp.
About 2 years ago, my then girlfriend, who is a medical doctor (read: likely not an idiot), had ChatGPT plan a trip for us in a foreign country (naturally I had to nip it in the bud). And this was 2 years ago, when accuracy was way worse.
You would be surprised what normies are willing to delegate to AI.
I often won't look at it's route. If it seems like something's wrong along the way of course I'll double check.
Those folk driving off bridges and into ravines is scary as hell, it implies there's people driving around that don't look further than their car bonnet whilst driving (-+5 metres).
I haven't seen the livestream, but I just heard that they intend to have their AI automatically change your passwords on websites if it considers them insecure, which sounds to me like the worst idea for AI so far, and that bar is high.
The better use case would be to make AI cancel that damn subscription that lets you jump through 20 dark pattern questions and then tells you to call customer support.
And if something breaks, and something will break - it's Software+Apple, their support will talk to you for 3 hours very professionally, giving you the scenic route of everything IT support has done in last 300 years and then they will schedule another call, apparently with an expert, on which you will be told to reboot your devices (yeah, all of them), and next stop will be asking you to reinstall your devices clean, of course they will remind you to backup data and how iCloud plans can help. After all that you will be asked to go to a support centre and drop your laptop there (that is, if your device is still under warranty).
> The company reiterated that Apple Intelligence relies on on-device processing and Private Cloud Compute, with a promise that user data is only used to execute the immediate request and is not accessible to Apple or third parties. Apple added that outside experts can verify those privacy guarantees "at any time."
Yes, the "Apple needs to look at your data to do this, but we don't have any way to look at the data if we wanted to". That's impossible, unless they open souce iOS and let people take control over their devices, and let people self host inference, so people can check that there is no network traffic. If it is as they say, they could let people host it without any downsides.
Apple's PCC is the best option for this kind of offload that exists.
However the PCC root keys are still signed by Apple which requires you to trust Apple and the laws in the jurisdiction Apple operates in.
Edit: for this update they seems to be running Gemini on Nvidia GPUs in Google's cloud[0]. How key management works for this part is unknown, but the standard setup for this is that Nvidia and Google would have keys too.
It does use the OHTTP relay[1] which makes it hard - maybe impossible - for Apple to hand over the keys for a particular person's data. Maybe that provides some additional protection in US courts against overreach.
Is this a problem for most people? Probably not - but it is something to be aware of.
I think Apple have made a great attempt to make this as safe and private as possible, but until we have a truly trustless E2E encrypted execution environment I don't see how compute offload technologies gets around this problem.
[0] > And to bring this model to production, we work with both Google and Nvidia to extend our Private Cloud Compute infrastructure to NVIDIA GPUs in Google’s cloud, while maintaining Apple’s unmatched privacy guarantees
I've been a paid subscriber to Claude for a couple of years, but lately I've been reaching for the free Gemini app on my Android Pixel 9 because it's so good at doing searches as part of its answers. The model feels fresh and up to date. Whether Apple can incorporate that search is an open question
A lot of people are missing that Google is light years ahead in terms of edge AI. They've been going on about it even before the GPT-craze. Pixel phones have had live captions (on edge transcriber) for a while.
Things changed since Apple Intelligence but I was hoping there’ll be more things like live captions and what-not than chatbot use cases. I feel pixel is also moving towards that and abandoning the old way unfortunately.
Only if you consider Google Image Search and Google Nano Banana to be "the same thing" since they both produce an image based on text input!
Similarly, Google Translate's millions of lines of hand-rolled code has been entirely superseded by LLMs that do a vastly better job.
The LLM-based AI assistants are based on a wildly different technology stack with very different capabilities compared to the legacy "if-then-else" logic programming that Siri was based on.
Was Google Translate millions of lines of hand-rolled code? The Transformer architecture was invented for Google Translate, before it was used to build "LLMs".
>The company reiterated that Apple Intelligence relies on on-device processing and Private Cloud Compute, with a promise that user data is only used to execute the immediate request and is not accessible to Apple or third parties. Apple added that outside experts can verify those privacy guarantees "at any time."
> Apple added that outside experts can verify those privacy guarantees "at any time."
To my understanding, these "outside experts" have to go through a vetting process by Apple first. There are no publicly available audits of the infrastructure planned right now.
This move kind of reminds me of the original iPhone with google maps. You're competing with google, but you're using their infrastructure. Why wouldn't they just go with another provider like OpenAI or Anthropic?
To be fair, with the original iPhone they weren't competing. Google did backend web stuff well, so Apple partnered with them and made nice local apps that were fed by Google's cloud data. The same was true for the YouTube app.
After Google saw the iPhone (before the public), they pivoted their vision for Android (it was originally on blackberry-style hardware), and that's when the "thermonuclear war" started. Kind of interesting Steve Jobs would have showed the iPhone off like that, when something very similar happened with Bill Gates, which prompted the development of Windows (according to Jobs).
19 years later, it's probably time to be pragmatic again. If Apple isn't able to deliver on some of these AI integrations into the OS, they risk losing users to Android. If they have to pay someone for a model, they might as well choose the one they think is going to be best for their users. This keeps existing iPhone users on iPhone, and may pull over some Android users looking for the same features, but with better privacy. That seems like a win for Apple. To pay OpenAI instead of Google would just be spite at this point. Maybe well deserved, but the leadership has all changed over in the last two decades, so they'd be fighting old wars. Though I think they should still think twice before showing Google anything that hasn't yet been released.
Given that they had originally selected OpenAI for Siri, and that deal fell through, I would guess something about their relationship with OpenAI fell through. Maybe OpenAI wouldn't let Apple run their model on Apple's servers.
OpenAI or Anthropic are not anywhere as well funded as Google. Apple already has everyone in their pocket via the ecosystem, they just have to not crap the bed. They value stability over the competitive component here.
Google is almost closer to a conglomerate than a coherent horizontally integrated corporation. The individual parts of Google are like Fortune500 companies themselves, and tend to act in their own interest.
OpenAI and Anthropic, despite bombastic media coverage, are still frontier labs. Google won't go down under if Gemini doesn't sell enough anymore. Apple and Google are planning for the aftermath of the AI IPO craze, one way or another.
I’m not thrilled about any sort of Apple AI. I see it more as the convenience of platform lock in that ideally would be in the hands of all serious AI contenders but we all know that’ll never happen.
Every mainstream product seems to have their own “SmarterChild on steroids” bolted on top (Gemini for Google, Rovo for jira, Copilot for Microsoft everything, etc).
I’ll still use the serious ones like ChatGPT/Claude as my main but I think these companies know that and are just trying to jump the bandwagon so they don’t look outdate. Either way, they can be surprisingly convenient and make up for UI/UX learning curves.
I used to wonder what "apps" might become in an "App Intent-first" world.
Bundles that provide data and capabilities to iOS and Siri? And perhaps libraries of UI components to display and interact with said data?
But then, if that works really well, and gets strong adoption, why ever open the app? What’s the point of having navigation flows inside an app? Could one make entire apps solely dedicated to providing a set of data, capabilities, and UI components to the system?
In that world, what drives user retention, for such apps? What even is an app? App engagement disappears as well.
And that’s not even diving into the use-case of Siri, say, planning a trip across five different apps (flights, hotel, restaurants, whatever) using just App Intents. If done well.
In that world, do most apps just become plugins, providers for Siri?
If you remember Windows Phone 8, it has already been tried and nobody (ahem, no company) wanted that.
No company is stupid enough to give up their content and infra and get none of the screen real estate.
I can see a parallel with hotels and OTAs, but in that case appearing on an OTA brings in sales. Showing $userA's content on $userB's screen won't earn any money from $company.
Well, there’ll always be room for people who want different UI/UX. Humans are too visual to ever move to “pure voice” and so we’ll inevitably have nice screens and thus UI preferences (technically you get preferences with voice too but it’s weirder)
Yes, but you can have this by providing surfaces as UI components to Siri.
Not sure if they do that (yet), but no reason an app couldn’t expose "Here’s what you can use to present data of shape X", or "here’s a UI for process doing y".
It feels like turning the common approach inside-out. But it works.
Edit: you could even imagine, in that world, apps that only expose surfaces, composable UI libraries, multi-step flows, declaring what they’re for, what kind of inputs they take, and what output they produce. Without ever owning any of the data (eg flights data, hotels inventory, booked trips, financial data, etc) or capabilities (eg book a flight).
That is very logical. Google's AI has rapidly become better at search and presents answers in a tolerable way.
If we disregard exploitation of the primary sources and focus purely on a use case, search summaries are it. Generative AI is completely useless and OpenAI and Anthropic will soon fail. The Segway is maybe fun for a day, then you ask yourself why the heck you aren't using a bike in the first place.
Everyone assumes this is apple paying Google to use their stuff. But maybe its the other way around. Maybe this is the google search engine deal 2.0. i.e. Apple agrees to sell their customers to Google for money. I know that right now its supposed to run privately and all that, but its still Google's voice telling you things, not to mention intepreting your instructions.
Google right now offers plenty of AI inference for free for everyone on the planet. They also offered their £20 AI pro package for free for over a year for all students. I use that and its a very generous offering. Some more free AI inference for Apple users (where you've now selected for high income folk, simply by virtue of being Apple users) would make plenty of sense, I imagine.
for those commenting on relying on a competitor for a "major" component of their product, look no further to how they source their displays from samsung all this time. even their silicon at one point! [1]
while it is now all about harness and "holding it right", their approach seem no different from other instances on hardware side. in terms of why not a more "leading" provider, look how the openai integration panned out so far. moreover, they have little to no initiative nor incentive to provide edge models or a hybrid solution so far. and anthropic's moat is not in audio.
It does, I agree. That said, they've published some Apple Research docs/papers & core documentation where they outline the architecture and how it works. Personally I think their approach is fascinating.
It is weird and disturbing that Apple has no native AI capability.
This is one of the most cash rich companies in the world and it has failed to have any position in the most critical technology development perhaps ever.
It's a clear signal that Apple became the most incredible operational/execution company under Tim Cook, but lost its innovation leadership.
You don't stay the most cash-rich company by chasing every expensive fad and they've been equally conservative with other "THIS WILL CHANGE EVERYTHING" tech fads such as Cryptocurrency and VR. I don't blame them for not rushing to light Billions a month on fire like the other big players; their play always seems to be to let things shake out and then deliver something refined and sophisticated.
There also doesn't seem like any real opportunity for them to Apple-ify this tech (any more than today's announcement). There's lots of rough edges and the underlying technology is fundamentally janky and extremely problematic in Apple's second differentiator of privacy.
I take it as a signal they don't see any of their value being provided by the models. They're strong point was never frontier technologies. It's always been the delivery of the technology.
They are a hardware company at heart, the do software as software is needed.
For building a competitive AI they'd have to hire the talent, which is expensive and then do a massive investment, which may still end up far behind the competition. (See there attempts with Siri)
Now they can pick the model they want and if time is right they can still build their own.
In the end they still want to sell devices. They aren't doing a search engine (while they could), they are not doing an LLM model, ...
>> They are a hardware company at heart, the do software as software is needed.
This has never been true, not since Steve Jobs returned.
The heart of Apple is software and hardware integration.
Hardware companies that do software just to prop up the hardware business do terrible software, and (no doubt the Apple haters gonna hate this) but Apple does - for the most part - amazing software.
>> which is expensive and then do a massive investment
Come back when most of AI model makers go belly up....
Apple largest acquisition in their entire history is three billion dollars (Beats), Sam OpenAI got nothing and Google is getting a measly one billion dollar refund for Gemini.
Innovation is Apple Silicon and the five ecosystems (Microsoft and Nvidia aren't sparking now for nothing), Innovation is being the last American vertical computer company left from the 1980's who has been able switch gears went necessary (the next gear shift will probably be memory)...
Well it's half-and-half, why did Apple struggle for so long with Siri and its pre-LLM era technology, during the time of AlphaGo and so forth, and then after Covid why didn't Apple pivot to something like their own version of Gemini?
But there are lots of differing possible reasons for this, and I think it is premature to conclude with any one in particular.
I have been using their on define AFM models for a year - for small models they are good. Their Secure Enclave server bases AFM model is good, but not in the same class as gemini 3.5 flash or deep seek v4 flash.
Really I don’t think this is a strong take at all. If anything this has positioned them extremely, extremely well for when the bubble bursts and they can go with the winner to provide reasonable capabilities.
It's looking like an incredible move so far. They will have wasted no money while the whole industry lit trillions on fire, and then at the end they can just rent a model for cheap and replace it with their own or whatever is cheapest at any time.
The users really aren't Gemini users, they don't care what the model is behind the scenes.
They should have built a separate phone around Apple Intelligence. That is a no brainer. Jobs would have done the same. That is not going to be easy, who knows about user experiences better than Apple does (or atleast used to). This is underwhelming from Apple.
Google sold Apple the ability to run certain Gemini models on Apple's data-center hardware, using Google's orchestration layer. Apple hooks into that not dissimilar from an API-provider, and then builds everything upstream.
Meaning the system prompt(s), harness, entry and exit points, and skills. So the product is still "Siri AI", because of all the stuff that takes it from a raw infrastructure concern upon up into a "product" is Apple's responsibility.
Google are "okay with that" because Apple pays them $1B a year, per press reports, to be.
Any business can do this now actually - you just need to lease/rent hardware through a Google partner and you can run an Nvidia based server in your own datacenter running (supposedly) the latest full Gemini models.
I don't think you conceptualize Google's game plan. all these companies care about is b2b contracts so they can inflate their balance sheets because when it's digital, it doesn't have to actually exist for it to "make money"
Let us hope the EU forces Apple to allow the end-user to choose the external model: Wouldn’t it be amazing having privacy first local models calling out via a welldefined open protocol to a model of your choice: Claude, Grok, DeepSeek?
Sounds like OS architecture done right - screw the kickback business model.
That's basically what we built at Tinfoil. We run open source models inside secure enclaves (also using Intel TDX/AMD SEV-SNP + NVIDIA Confidential Computing). All the code running inside the enclave is open source and the client SDKs (also open source) automatically verify that the pinned source code matches the runtime attestation. The protocol used is TLS (terminates in the enclave) + HPKE keys generated inside the enclave on boot. Docs walk you through the verification process: https://docs.tinfoil.sh/verification/verification-in-tinfoil
Of course, we can't support Claude or Grok as they are closed source, but there is no incentive for companies that need your data to train the next generation of models to allow for private inference. One day...
Funny, OP wants them to force a feature via regulation, and regulation is the reason they won’t even deliver the feature in question. Death by regulation.
Let's actually not hope that, and let's not indulge the EU regulators' fantasies that they get to dictate the product design of products from non-EU countries.
Force Apple to support all kinds of arbitrary models? That's a comically bad idea.
Why would it be ok that a monopoly business prevents free competition and consumer choice by only allowing certain or a single model provider that likely gives them kickback via some opaque business deal?
I tried the on-device image input model with an app I am building. It's not very good at world-knowledge recall, but can describe the submitted image well enough.
Sort of expected this at the first attempt. Use their existing partnership for Google being the default search with Google and leverage Google's models.
Sorry to be off topic, but I have a question: has anyone installed the latest beta iOS and macOS, and if so what is the current status of Gemini integration?
once you update to the latest iOS, there's an option to join the waitlist. No indication of timing, but until that moves forward Siri is what you get today, as far as I can tell.
Google apps are the most downloaded apps in the Apple App Store already. This reminds me of the original Apple Maps, which was just a front end for Google Maps.
If you pay for Gemini, then it is good. I recently used Gemini Ultra for a month and the gemini models are very good (and of course, you get a lot of Claude Opus tokens to use through the same plan).
I also pay for Proton's Lumo+ private chat and for what it is it is also good.
The free plans from all the providers are bad, which is fare enough.
I use Apple devices and I expect to be paying for Gemini tokens after the integration.
Just an interesting thought on the very different philosophical approaches. Let's imagine that Android had a terrible assistant, so bad that even most Google Fanbois admitted it was pretty bad. Apple became a leader in AI. Google approaches them to license the Siri model for Android. Would Apple have ever done that?
“Now, we are collaborating with Google and NVIDIA to run new Apple Intelligence workloads on Google Cloud, extending our industry-leading PCC privacy commitments to third-party data centers for the first time.”
It’s the DMA regulation that forces Apple to give the same access as they have to other AI chat apps.
Once it leaves the device Apple does not know what those other ai chat apps will do with the gathered data.
> Siri AI is private by design and deeply integrated across Apple’s platforms using on-device processing and Private Cloud Compute, which extends the privacy and security of iPhone into the cloud. However, under EU regulators’ extreme interpretation of the DMA, Apple would have to give any virtual assistant direct access to users’ private data — and the ability to directly control other installed applications — as soon as Siri AI is made available in the EU, without the essential protections necessary to keep users and their data safe.
Apple loves to play dumb about this stuff. The EU imposes a pretty straightforward regulation regarding equality of access. Apple seems to come up with all sorts of "solutions" to this "problem", and each one never amounts to true equality of access. They could easily just allow users to decide "Do you want to give this app unfettered access to all your device data, including other apps' data?". Let users decide. 99% of Apple users in the EU will probably click "no". I'm sure they'll make the user warnings scary enough to ward off anybody who doesn't know what's going on.
There are 2 potential outcomes: either the sky really does fall, and there's a meaningful uptick in bad things happening to iPhone users, in which Apple can easily point the finger at the EC and say "they made us do this". Apple looks like the good guys who put up a good fight for their users, but ultimately their hands were tied, and they'll probably get the revisions to EU law they're so desperately fighting for.
The other possibility is that the sky does not fall, and Apple looks both silly and malicious at the same time for ever having suggested that it would, which was clearly in bad faith.
Clearly, Apple cannot afford scenario #2, so I think they will probably never give their users the actual freedom that the MDA requires them to. They will just exit Europe entirely before allowing that to happen.
> It’s the DMA regulation that forces Apple to give the same access as they have to other AI chat apps.
But why can Tesla ship Grok to their cars in the EU without any problems? Why aren't they required to let me choose between Grok, OpenAI etc or even a custom endpoint?
Since it's the user's device, not Apple's, EU correctly "interprets" this as the user has the right to do whatever they please, including installing third-party chat apps.
Apple are just bulshitters when it comes to actual users, and not their corporate definition of a user.
Unless Apple proves otherwise I'm more inclined to believe they're either 1. Using this to try and shape the DMA in their own interest (definitely not their users' interest) or 2. Doing something with the data that would not be allowed in the EU (also not in their users' interest at all) or both.
They say it's because the EU's DMA would require them to open up device data to third-party assistants, and they'd no longer be able to guarantee user privacy.
I don't see what the issue is. The user could then select Apple (or Mistral) for strong privacy or another provider for customers that don't care.
I primarily want Apple to provide extension points so that I can select my own provider, just as I can choose where to host my mail or install another app as my instant messenger.
Sure, I could install another provider's app, but it wouldn't have the same integrations, similar to how an instant messaging app would be less useful if notifications were limited to iMessage.
Nothing in the law requires the pop up. It definitely doesn’t require the obnoxious bullshit that most companies put up (aka the dark pattern to get you to agree to every unreasonable part of their terms just to read the page).
The alternative would be to just stop invasive tracking and add the cookie when it’s actually needed.
The smell is that Apple doesn't want to give the same level of access to third-party AI assistants that Siri will get.
For what it's worth, Apple claimed they proposed an "equivalent access" framework with some kind of "trusted agent framework" approach, but that it was shot down by the EU. I suspect it was way more inconvenient for third-party developers than Apple lets on.
This has been the case for quite a while. Like Reminders - You can't replace the phrase "Siri, remind me to ___" with a third party app. I'm surprised the EU lets them ship Reminders there.
what i think why they are too much relying on Google is coz they are way much towards making models open source and launching more much better models to public as if in future apple part way from google they might still have much better models to rely on and if we see the history google has been partnered with Apple since the launch of first Iphone
What a blunder, they resisted AI for like 2 years when it was all the buzz, and now when the bubble is about to bust and every user has AI fatigue they decide to finally dip into the fad?
Before it was as if avoiding AI was a conscious design decision, and if there was an AI crash, Apple would be the only survivor left. Now it feels like they weren't in on the meme out of incompetence and are now late to the party.
No one can know what Jobs' stance would be, but I like to think he would be anti-slop
I agree that many AI businesses will go bust and they deserve it, but the tech is good.
I can recommend my own layered approach, using the lowest capability models that get stuff done:
1. I maximally use local models like gemma4:26b-a4b-it-qat for everything that works with this free option.
2. I like paying for inexpensive APIs for mid-tier models like deepseek v4 flash, gcp-5-mini, gemini-2-flash for things that option 1. fails at. This option is almost free.
3. Pay for more expensive APIs like deepseek v4 pro, gemini 3.5 flash, etc. This option is not too expensive.
4. If all else fails on a class of tasks, then pay for awesomeness of Claude Opus. $$ expensive, I try not to use unless absolutely necessary.
I think developers and companies that just cram everything into Claude Opus are unprofessional.
LLMs we all agreed were amazing back in 2023-2024.
What's happening now with AI is more of a corporate phenomenon quite removed from the actual tech.
Yes LLMs are useful, but replacing customer support with an LLM that gives user accounts away, or calling LLMs on a loop where the bottleneck is your checkbook and calling it AGI, those are phenomenons that are separate from LLMs.
When people say the AI bubble is about to bust, I don't think anybody means that "the use of AI is going to go away." AI is absurdly useful. I think what people mean is "the valuations of these companies will have to snap to a reality that is actually attached to their market value."
Exactly, small edge models is the future, highly personal experiences, and not these massive models that the cloud providers currently shove down our throats. While massive models are useful, these massive platforms are about to burst out of their promises. All while we’re supper happy with tiny 4b up to 12b models working amazing for all these “omg ai thinks” daily tasks.
I honestly don’t understand how anyone can believe that Apple is limiting user options for privacy reasons, rather than trying to maintain an unfair advantage over other vendors.
I’m not saying people who hold this view are being dishonest at all. But sometimes, to me, brands like Harley-Davidson or Apple seem closer to a cult than to a typical corporation.
Because privacy is actually a big feature, many people are skeptical about AI and the big model providers and don't trust them. They're less skeptical about on-device AI and so Apple is pushing that and making privacy a core feature of their online offering as well.
I probably wouldn't use it without that. It's one thing sending my shitty code to be trained on, and another thing entirely to give these companies access to my personal life and information.
When you have to image a highly irrational reason to explain why groups of people do the things they do, there's a decent chance you just don't understand their perspective. They may be acting reasonably rationally from their own perspective. (As you said yourself: "I honestly don't understand...")
Apple could have done something like bedrock and used a SOTA model but instead they are fiddling with local models or whatever.
Also I have seen that Apple has some strange lust towards image generation as if that's what people really want. I have this slop image generation thing on my phone and it is useless.
Here's what I want: natural language interaction to achieve complex workflows in iPhone. Example: find the cheapest way to go from A to B and book it using the Deutsche Bahn Train app.
I don't disagree that local Gemini is better but if you've tried it in iPhone, it is slow, hallucinates and overheats the phone. For anything slightly non-trivial like the workflow example I gave, I think it will be close to useless.
Very Apple-ish approach to AI catch up: wrap an external tool in a privacy architecture, embed into the OS and productize the orchestration layer.
It will be interesting to see if the Private Cloud Compute + on-device routing can make third-party model capabilities feel like a first-party system without leaking user context to the model provider.
If Apple handles the Google-Apple boundary right, this will be an elegant move on their part, otherwise it will feel like Apple Intelligence with a just a privacy-polished frontend for Gemini.
> wrap an external tool in a privacy architecture
Wouldn't it be more accurate to call Apple's architecture data protection rather than privacy? As an European citizen in a post Snowden world I would be surprised if any of my data on Apple services was actually kept private from the US government, and Apple certainly wants to own a lot of data/metadata about you. Gotta have Siri listening for carplay and so on. I would aboslutely trust Apple not to sell my data as a commodity though.
> If Apple handles the Google-Apple boundary right, this will be an elegant move on their part, otherwise it will feel like Apple Intelligence with a just a privacy-polished frontend for Gemini.
I'd say this is spot on. At least if what Microsoft is doing with Copilot Cowork is anything to go by. Cowork is not a privacy-polished as much as it's an Enterprise compliant polish to make Opus 4.8 run "safely" in your enterprise organisation. So far Microsoft is winning the AI war in non-tech enterprise with this, especially here in the EU. If Apple manages to do this for the private market that will be great for them.
I'm not personally sold on what an AI should do on my phone though. I use a lot of AI professionally, but I haven't even turned on Bixby or whatever the Samsung AI is called.
From an EU perspective, Microsoft is doing data protection, Apple is doing data privacy.
Microsoft's approach to data is basically "we promise nobody else but you and your government can access it, we can but we pinky swear we won't." This promise is mostly enforced at the legal layer and through legal consequences, not technical safeguards. If they think they can get away with it (or are forced to get away with it by the US government), there's nothing stopping them from using your data in whatever way they want.
When they can, Apple designs their systems so that they physically don't even have the capability to use your data, even if it's processed on their own servers. They're not privacy maximalists like Signal is, they care more about user experience, but they do aim for the highest level of privacy you can get while still having a good experience, and when they do need to make sacrifices, they typically let you opt into the privacy features if you really want to.
I'm far more inclined to believe that Microsoft is secretly (or not so secretly) collaborating with the US government than that Apple is.
8 replies →
> Wouldn't it be more accurate to call Apple's architecture data protection rather than privacy? As an European citizen in a post Snowden world I would be surprised if any of my data on Apple services was actually kept private from the US government, and Apple certainly wants to own a lot of data/metadata about you.
Your conception doesn’t seem to match PCC at all. The whole point of it is that nobody can access the data, not even the people running the servers.
https://security.apple.com/blog/private-cloud-compute/
27 replies →
> I use a lot of AI professionally, but I haven't even turned on Bixby or whatever the Samsung AI is called.
I know this was just a small aside, but man do I hate Bixby and other phone AIs. They are so frustratingly difficult to turn off, and turning them on accidentally is as simply as holding the wrong button for a few seconds, such as when your phone is in your pocket. Very frustrating design.
3 replies →
> I would be surprised if any of my data on Apple services was actually kept private from the US government
Outside of law enforcement having a warrant, Apple's efforts against CSAM, or their Chinese data centers, I've not heard of Apple doing any of what you assume in a post-Snowden world. iMessage is supposed to be end to end encrypted, and there was a few years ago that whole scandal where Apple wouldn't unlock a literal terrorists cell phone for the FBI.
The FBI had to reach out to... a third party to unlock the phone (I forget the name of the firm that did it - Cellebrite maybe?) for them, what's funny is they spent a lot of money on it, when the rest of the world pointed out that the very specific iOS version in question had known vulnerabilities they could have found online for free (or cheaper?).
Is there a meaningful Google-Apple boundary in operation?
They are buying the right to distill their own Gemini models and run them in their data centres (or at least data centres they control); unless I am missing something, this isn't going to be infrastructure that Google has operational control over.
If Apple is running the inference from Apple iPhones and Apple data centers then Apple has operational control. Google’s influence ends the moment they hand the weights over to Apple.
39 replies →
As someone who doesn't use Android, they showed a lot of integration into the apps, which I think is where the real magic happens, and it's not something I can do with any 3rd party chatbots today (that I'm aware of). I also don't know that I would trust the other 3rd parties with the access required to pull it off.
If anything, having Shortcuts built-in at the OS level was a very long play that’s going to pay off.
1 reply →
I'm interested in how it feels to use: whether there is any context leaking, as you mentioned, if it introduces latency, and whether there are any pricing implications? I know they weighed a variety of factors, including the smaller models, but cost had to be a big concern, too... I feel like Google is the only provider giving away so much AI inference for free.
This will further blur the picture about when and how consumers / employees are supposed to pay for AI services. For example, they showed consumer rather than coding tasks, but could you select five files and ask Siri to write a Python script or a small app? Will enterprises just disable Siri AI functionality, or will they be able to route it through their own AI auditing and providers?
Oh yes Apple "wraps" their AI in a "privacy architecture" -- you can't use Carplay unless you turn it (Siri) on.
Yes, but Siri can be turned off from invocation without turning off CarPlay. You can disable the side button and Hey Siri while leaving Siri "on."
2 replies →
How do you expect to use a nav system while driving without voice control?
21 replies →
“Sorry, I don’t know where you are”
> just a privacy-polished frontend for Gemini
I'd use this.
I'd rather have strong privacy guarantees, but this is still good.
The most Apple-ish thing would be to produce a great platform, enable third parties to do something on top of it, and take a cut.
I REALLY wish they'd do that with voice assistants.
All I know is that Siri is a terrible user experience.
Consider that they described the development process as taking Siri down to bare metal and rebuilding on a new architecture. I don’t think our previous Siri experience will be particularly relevant.
2 replies →
This is post Jobs era typical Apple.
[flagged]
[dead]
[flagged]
[flagged]
I absolutely adore the historical revisionism that apple cares about privacy.
Run your router through a linux laptop as a proxy so you can capture traffic, connect any apple device to your router, and see the vasts amount of data your device sends to apple.
Apple DGAF about privacy, they want your data as much as anyone else, their only thing is that they should be the only ones to get it and then other people have to pay them for it, rather than your device sending the data to the 3d party directly.
And if you think your data is secure, reminder that The Fappening was all done targeting apple devices.
Apple added e2ee and created the most complete end to end encrypted cloud ecosystem to prevent that from happening again.
6 replies →
I would love to learn more about what's actually powering Apple Intelligence now. Are they using flagship Gemini models behind their own prompts? Fine-tuning? Pre-training their own models based on Gemini?
Is there a meaningful distinction between the Gemini-powered models and Apple Foundation Models? Does that distinction vary for on-device vs hosted models? Are some models running on Apple's Private Cloud Compute and others running on Google iron?
Edit: they elaborated significantly in a "keynote tech-talk": [0]
According to Apple, there are five models:
On-Device
- AFM Core: Dense architecture; the standard next-gen on-device model
- AFM Core Advanced: Sparse architecture, natively multimodal; enables features like image understanding and expressive voices
Private Cloud Compute
- AFM Cloud: Workhorse server model optimized for latency and cost
- AFM Cloud Image: Image generation and editing
- AFM Cloud Pro: Most capable model, Gemini frontier-level quality, for complex reasoning and agentic tasks; runs on NVIDIA GPUs in Google's cloud under Apple's PCC privacy guarantees
Everything excluding Cloud Pro are custom models running on Apple Silicon, "refined" using Google Gemini. About Cloud Pro, they say "this is our most capable model with quality similar to Gemini frontier models." So I might read between the lines and say this is a wrapped Gemini.
> what's actually powering Apple Intelligence now.
It's a 3B Apple Foundation model.
https://machinelearning.apple.com/research/introducing-apple...
If you've got a mac, you can use this to play around with it:
https://apfel.franzai.com/
I think what they mean by “now” is the stuff announced today.
It's more complicated than that (see my edit above).
Thanks for that. I like being able to pipe output into a local LLM to get an explanation of the output.
That’s really neat. I wonder if that model that shipped recently with Chrome is also accessible similarly?
> Are they using flagship Gemini models behind their own prompts? Fine-tuning? Pre-training their own models based on Gemini?
Easiest way to tell, how much dancing around did we listen to and how many diagrams did we have to look at? If they had their own tech we wouldn't be looking at diagrams we'd just be getting told Siri AI, it's private, it's powerful, here's what it can do. Instead we had 10 minutes talking around the tech and this diagram [1] which is a signal that it's a bunch of other peoples stuff cobbled and wrapped together.
[1]:https://www.apple.com/newsroom/images/2026/06/apple-introduc...
Local is probably similar to Gemma e4b you can get right now on Google Edge Gallery (the ios and Android app). Guessing that the more powerful version that will only work on the 12gb ram devices will be something unreleased that is similar but a bit larger
Google also awhile back announced being able to run full Gemini by leasing / renting hardware in your own datacenters so companies can train or access data without needing to send things to their datacenters. Nvidia based. Guessing Private Compute might just be Apple leasing a ton of those?
Apple Private Cloud Compute is running on M2/M3 Ultra. I'm not sure if Gemini Flash can fit in that amount of RAM.
A larger context window would be nice, the apple model on devices now is almost too small to do cool stuff with
Am I reading this correctly? Their chosen cloud providers run the PCC stack on their hardware, so the compute provider is responsible for ensuring the privacy guarantees? I assume that would add to the potential security surface area.
Intel and Nvidia are responsible for enforcing their privacy features. The cloud operator (Google in this case) has no access to any data.
Yes, that seems to be the case, and is an evolution/deviation of the original PCC model, which relied on Apple Silicon exclusively.
1 reply →
> Everything excluding Cloud Pro are custom models running on Apple Silicon, "refined" using Google Gemini
What could "refined" mean here?
I was thinking distilled?
Always appreciate people answering their own questions in great detail!
[flagged]
Gemini (at least public free version) hallucinates way too much. If it's like that, it can go very badly for Apple.
I used Gemini exclusively via the API but downloaded the app last week for something. Even on max settings, it is ridiculously nerfed!
1 reply →
The public version of Gemini is ridiculous. At least half their search "answers" are just wrong. If you then start a follow up chat the answers change but usually still half wrong.
Search would be better without the added AI hallucinations above it. If I want an AI answer I'll go and ask Claude, the quality difference is huge.
6 replies →
It's strange to me that Apple would choose to disadvantage themselves by selecting Google as their provider as opposed to, say, Anthropic or even OpenAI. Doesn't this mean they'll struggle more to differentiate themselves from the assistant on Android phones? Thinking more cynically, couldn't Google, if they wanted, feed Apple an inferior version of Gemini, ensuring they stay ahead?
As the consumer, this just sucks because it means no matter which phone platform you choose, you're getting the same thing underneath, and there's no way to avoid it (besides not using an assistant entirely, which I recognize a lot of people do, myself included).
I'd wager that for 99.9% of "Apple Intelligence" tasks, Google's models perform just as well as other frontier labs. Google also has done more work on getting LLMs running on edge devices compared to anthropic and openAI.
The source also says > The new architecture centers on Apple Foundation Models co-developed with Google, which Apple says are adapted to run both on-device and on servers through its existing Private Cloud Compute infrastructure
Which could mean Google and Apple have trained some custom models, probably the on-device ones, specifically tailored towards Apple's hardware.
That's not the point of OP. The point of OP is that Google is a direct competitor of Apple for phones / OS, so them giving the "key of the house" to Google is risky
14 replies →
I see it differently… Apple has chosen to treat the model as a commodity. By making Gemini an implementation detail, they leave the door open to swap it out for Anthropic or OpenAI without end-users even knowing or caring. So I think they're creating leverage in any future negotiation.
And at the same time they claim it can't be rolled out in Europe because the DMA would force them to allow selection of other AI providers? Which would not even be true if this was an Apple product built on the models of Google, just like the DMA does not force them to pick a different datacenter of office cleaning provider.
8 replies →
> I see it differently… Apple has chosen to treat the model as a commodity
It didn't work out well for Yahoo.com. It turned out that Google Search was the value and yahoo.com just skin around it. It might be the same for Apple. Gemini is the valuable part, what particular device you use it on matters less.
3 replies →
In time Apple will swap it out for a in house model like many other things they’re swapped out in the last 25 years, Apple appears to be a company that doesn’t waste money and seem to execute long range projects if necessary I don’t think the Google models will be there for long. I think they will be swapped out when the M series GPUs get to the performance level they want.
It’s a bit like when Steve Jobs turned down acquiring Dropbox telling them they’re just a feature, not a product
2 replies →
I thought this seemed significant at the time: https://github.com/ml-explore/mlx/pull/1983
That’s exactly what a model should be: Implementation detail.
[flagged]
Maybe they're looking for stability and trust Google to be around longer than Antrophic or OpenAI when the storm starts.
Bingo. They see this as the future commodity it will be. Customers will choose AI providers much way they choose a car: taste, price, a few other factors.
And to your point, Google has a massive balance sheet, produces their own AI chips, and is not going anywhere anytime soon.
2 replies →
Google was the right choice stability and it only cost Apple $1 billion per year that’s pretty much of a no-brainer, and with Apple’s history, they probably will use Gemini for as long as they need it and then use their own model in time.
And Apple already has a $20B search partnership with Google they can build on.
1 reply →
I’ll just put this link here: https://killedbygoogle.com/
1 reply →
I may be jaded, but I do not trust Google for product offering stability. Obviously, Apple is a way bigger fish.
1 reply →
> It's strange to me that Apple would choose to disadvantage themselves by selecting Google as their provider as opposed to, say, Anthropic or even OpenAI. Doesn't this mean they'll struggle more to differentiate themselves from the assistant on Android phones? Thinking more cynically, couldn't Google, if they wanted, feed Apple an inferior version of Gemini, ensuring they stay ahead?
Is it really all that difference from Apple defaulting to Google's search engine?
> Is it really all that difference from Apple defaulting to Google's search engine?
Not really, because the business model isn't there (at least not in this iteration.
1. The models are Apple models, co-developed with Google. They are not white-label Gemini.
2. There's not currently a Google failover or UX
3. Because of that, there's no user monetization to share.
Apple does have a ChatGPT integration, with failover UX, and with a suspected revenue share deal. However, one could see this deal in a precarious situation, since at the time it started it was expected Apple would not focus much at all on a model capable with world knowledge.
Depends on which way the money is flowing. Google pays Apple for default search engine. Is Google paying Apple for using Gemini? That feels like a much heavier investment if they are
Yes, Google can do that just like Intel, Samsung, Nvidia or Qualcomm yes Google can drag their feet, we know in the end it will all lead to tears and then they will separate.
At the time Apple made this decision there wasn’t as strong of a difference in model quality between Google, anthropic, and OpenAI as there is now. Also Apple definitely burnt some bridges with OpenAI on the agreement they made together a year earlier.
> wasn’t as strong of a difference in model quality between Google, anthropic, and OpenAI as there is now
What's the difference now? I would guess 9/10 people here would have a very hard time telling the models apart in a blind taste test.
2 replies →
OpenAI and Anthropic don't make small models. Google happens to already have a billion devices that would benefit from small models, so they made one. Google basically gets 1 billion per year for free*.
Google has very good small models which can run locally on a phone - Gemma4.
OpenAI/Anthropic have nothing in this segment.
Did Apple indicate they are using local models?
2 replies →
Google was likely the only lab grown up enough that could handle Apple's volume and requests.
Apple was not going to hand over the keys to AI to just anyone.
Apple is a Fortune 5 company with a brand value alone worth more than any of these AI labs besides Google.
There's too much at stake for them to not play it safe. There's almost nothing to gain taking a risk
> Apple's volume
Yeah, I think Apple's volume made Google the only choice. And even then, Google was buying more DC capacity last week.
Gemini is better than either at multi modal, google also has their tensor processor stuff with ridiculously high T/s output they need for acceptable UX
Only Google was willing to do what Apple wanted for a reasonable price.
They licensed Gemini and Google infrastructure not just for use, but to accelerate the creation of the three independent Apple Foundation Models announced today:
- AFM Core
- AFM Core Advanced
- AFM Cloud
Google also worked to be able to host AFM Cloud on their infrastructure per Apple's private cloud compute architecture, including some form of independent third party review/audit.
I suspect the only two organizations with both the model and the infrastructure needed for Apple were Google and xAI - and I'm not sure Apple would touch Grok with a ten foot pole, even if xAI were willing to let it be used for training.
Google will probably eventually pay Apple to be the assistant, a la search.
Google pays to be the default search because they make more from selling ads on those searches than they are paying for the search.
I don't see the same thing here. Google isn't making any money from being the assistant in Apple, so why would they pay to be it?
4 replies →
Maybe. Search ads likely make Google more money than they pay apple. For AI, Google currently loses money. If they eventually make money via ads, then sure. Else, apple will have to pay them
1 reply →
Search has ads but Siri doesn't. And when Apple puts ads in Siri they won't be Google ads. I don't think Google benefits from this deal enough to be worth paying.
Eventually? I bet they already are.
3 replies →
You think Google cares.
Google doesn't care they do not control iOS. Google cares to have their products everywhere. For many years Google apps were better on iOS than on Android, because on Android they just preinstall them while on iOS they need to be installed explicitly.
Android is not a Google product, it is just a tool for Google to collect data. If they manage to collect data via Apple Intelligence they are going to do it. Regardless of what Apple marketing says.
It may be a clever move. By using the same models as android (contractually?), they can compete on the user experience which they typically handle better than android phone providers.
> or even OpenAI
Apple originally partnered with OpenAI. We won’t know all the details for some time, but given OpenAI’s penchant for drama (they started leaking that they might sue Apple [1]), it seems fair to sideline them as a long-term partner.
[1] https://www.macrumors.com/2026/05/14/openai-considering-lega...
It's not clear that there's any sort of durable advantage to the provider here -- in fact, oAI started as the partner with apple a couple of years ago, and is reportedly unhappy with the outcomes.
The hard part is not distilling a frontier model down into a specific use case when you have hundreds of millions of users, the hard part is (apparently) re-architecting your mobile OS to work with such a model rather than fight it. Those architectural benefits accrue to apple, as will future datasets and expertise, and the benefit of having some distillation working 24/7 on prem.
Anyway, where I think you're going to be grumpy in two years is that switching the underlying model is going to require a jailbreak, and that you wish they'd made the os much more deeply open for agentic interaction, not that it's gemini - it's just not the valuable part of the story for Apple or for users.
The differentiation is the integration, not the model itself which is mostly fungible. And afaik Apple is running these models on their own compute, so I don't think google can pull a bait and switch.
> It's strange to me that Apple would choose to disadvantage themselves
How exactly are they disadvantaging themselves? Perhaps expand on that opinion a bit, and the data/assumptions you're making in forming it.
My naive assumption is that they're going to do what everyone is doing: make tooling that lets you swap in any model.
I don't think it's fair to assume incompetence, on their part. I think it's much safer to assume they're doing what's best for them, and it's very clear to them what's best (money, support, etc).
> My naive assumption is that they're going to do what everyone is doing: make tooling that lets you swap in any model.
They are a ways away from that for Siri, as they can't guarantee third party tooling meets their security and privacy requirements. Meeting those security and privacy requirements also makes it harder for a third party to monetize their investment or ongoing use of infrastructure.
But I suspect you will see integration in other areas, such as image generation.
1 reply →
> It's strange to me that Apple would choose to disadvantage themselves by selecting Google as their provider as opposed to, say, Anthropic or even OpenAI.
The models are quickly converging to similar capabilities with particular ones being better at a particular task until the next release cycle.
> Thinking more cynically, couldn't Google, if they wanted, feed Apple an inferior version of Gemini, ensuring they stay ahead?
Yeah, that's not really how things work at this scale.
Differentiating themselves can only be a bad thing for Apple. Google is way ahead of Apple in AI and likely always will be.
Both Apple and Google end up advantaged by this relationship: Apple gets the same technology as Android, meaning there is no competitive front opening up. Google gets eyeballs on almost the entire smartphone market.
Apple can continue to differentiate the iPhone from Android in all the ways that they were doing before.
The agent harness matters just as much as the AI model. Using Hermes or OpenClaw feels like night vs day when using OpenAI’s apps even when using the same exact model.
You can even see difference in agent harnesses using the same model in the same company if you compare Gemini CLI with AntiGrav. They are different experiences.
I’m pretty sure Apple’s agent harness will be drastically different from Google’s even with the same model
Maybe openai wasn’t up to the level of customization and privacy they needed
Also openai and Jonny Ive (love from) are cooking some device — may be personal
Anthropic or OpenAI have no foothold into the mobile market. Google has integrated a shitton of AI functionality into their latest Pixel phones. That’s what would scare me if I was Apple and worried that if AI prevails this could steal some market share from me. The other two are irrelevant in this context.
I'm speculating, but it's likely that Google is the only provider who is willing to adhere to the inference compute requirements that Apple sets out for their foundation models. They are, after all, the only provider that will let you host their FMs in your data center.
> Doesn't this mean they'll struggle more to differentiate themselves from the assistant on Android phones
Realistically, does it matter? Most people aren't going to switch phone ecosystems over the assistant available on their phone's OS.
They tried with OpenAI and that deal fell apart. My hunch is that they're considering their own device play given they brought Jony Ive on board.
Anthropic doesn't have the spare compute laying around to do this deal. Even they're buying compute from Google.
What part of using Gemini do you think is a disadvantage?
Also important to remember how immature OpenAI and Anthropic are as companies. It would be a huge technical, legal, and reputational risk to commit to using them.
Didn't they famously have a search deal with Google (that they were also ultimately fined for - in the EU at least)? So there's definitely precedent with Google as a "partner".
A functioning FTC would not allow this. Insane that there will be no competition for integrated smartphone AI because the existing duopoly at the OS level has agreed to team up.
Or... because the existing duopoly of frontier/SOTA models (Anthropic and OpenAI) agreed to team up and not licence their weights.
Anthropic would never provide weight to anyone for local hosting
Google are the only one of the big three who can tick the boxes on being multimodal, price / performance and having Apple-level of compute available
Didn’t Apple talked to Anthropic as their first choice, but they couldn’t agree on an amount, almost similar to BeOS and Next remember them?
We don't really know, we just saw a leak when the deal fell through. I suspect Apple wouldn't reach out to just one AI company at a time.
You can be pretty sure that Google will be around us in a two year timeframe. Can't say the same neither about Anthropic or OpenAI.
Anthropic and OpenAI are stuck on slower and more expensive nvidia hardware. It doesn’t scale like googles TPUs.
Google's TPU and Groq's LPU are the only real, commercially viable ways to provide all the compute required for the inference people want.
Just having an ungodly amount of capex by blanketing the Midwest with datacenters full of GPUs is a disaster in slow motion.
They probably started working with gooogle before claude started moving ahead or was on the radar
I’m pretty sure both Anthropic and OpenAI don’t have the compute capacity for Apple users
>couldn't Google, if they wanted, feed Apple an inferior version of Gemini, ensuring they stay ahead?
Why are Apple people like this lmao. Yeah ofc they could, but they won't because 2 businesses as large as them have a deal it's usually honored.
Couldn't anybody in Apple's supply chain sabotage them? Contrary to popular belief, I don't think Google is in direct competition with Apple at all...Google don't make a mobile operating system. And they certainly don't make much hardware (especially that people should buy).
TPUs
If I can't even trust the results given by ChatGPT and Claude at their highest level of reasoning in my daily life and work, would I be willing to use Siri AI to handle the important scenarios depicted in the livestream?
The idea that someone would go on a multi hour car trip based on an AI-generated route without manually checking where they're actually going is quite funny to me. Some older HN members will remember when the first stories a la 'the GPS told me to drive down that ravine' started coming out. I'm looking forward to "my AI assistant routed me to Kangiqiniq, Canada and border security detained me when I couldn't produce a passport".
I sometimes think about the time I was on a road trip with a friend, and for what was generally a 3 hour trip to the mountain, was 5 hours for my friend driving.
Midway through the trip I was suspicious of the duration, traffic was fine. He was adamant he knew where he was going.
I pulled up google maps, and sure enough. 3 hours.
Turns out, his mapping app wasn’t aware of an offramp, so instead it wanted us to drive an extra hour, then do a u-turn and drive back to take an offramp.
He was blindly using Apple Maps
2 replies →
About 2 years ago, my then girlfriend, who is a medical doctor (read: likely not an idiot), had ChatGPT plan a trip for us in a foreign country (naturally I had to nip it in the bud). And this was 2 years ago, when accuracy was way worse.
You would be surprised what normies are willing to delegate to AI.
I often won't look at it's route. If it seems like something's wrong along the way of course I'll double check.
Those folk driving off bridges and into ravines is scary as hell, it implies there's people driving around that don't look further than their car bonnet whilst driving (-+5 metres).
I've used AI to plan entire 5 country itineraries multiple times and it's done better than I or any agent has
I haven't seen the livestream, but I just heard that they intend to have their AI automatically change your passwords on websites if it considers them insecure, which sounds to me like the worst idea for AI so far, and that bar is high.
The better use case would be to make AI cancel that damn subscription that lets you jump through 20 dark pattern questions and then tells you to call customer support.
3 replies →
You have to manually start the change, it doesn't do it continuously.
For the vast majority of services, even if this action fails and the wrong password is saved (!?) you're still just a "forgot password" click away.
4 replies →
And if something breaks, and something will break - it's Software+Apple, their support will talk to you for 3 hours very professionally, giving you the scenic route of everything IT support has done in last 300 years and then they will schedule another call, apparently with an expert, on which you will be told to reboot your devices (yeah, all of them), and next stop will be asking you to reinstall your devices clean, of course they will remind you to backup data and how iCloud plans can help. After all that you will be asked to go to a support centre and drop your laptop there (that is, if your device is still under warranty).
> if it considers them insecure
But that's all of them though?
Why do you assume Gemini is worse than those two? Especially not for code generation.
IMO if it is not absolutely net positive, it is very close to being that
> The company reiterated that Apple Intelligence relies on on-device processing and Private Cloud Compute, with a promise that user data is only used to execute the immediate request and is not accessible to Apple or third parties. Apple added that outside experts can verify those privacy guarantees "at any time."
Yes, the "Apple needs to look at your data to do this, but we don't have any way to look at the data if we wanted to". That's impossible, unless they open souce iOS and let people take control over their devices, and let people self host inference, so people can check that there is no network traffic. If it is as they say, they could let people host it without any downsides.
https://security.apple.com/blog/private-cloud-compute/
Apple's PCC is the best option for this kind of offload that exists.
However the PCC root keys are still signed by Apple which requires you to trust Apple and the laws in the jurisdiction Apple operates in.
Edit: for this update they seems to be running Gemini on Nvidia GPUs in Google's cloud[0]. How key management works for this part is unknown, but the standard setup for this is that Nvidia and Google would have keys too.
It does use the OHTTP relay[1] which makes it hard - maybe impossible - for Apple to hand over the keys for a particular person's data. Maybe that provides some additional protection in US courts against overreach.
Is this a problem for most people? Probably not - but it is something to be aware of.
I think Apple have made a great attempt to make this as safe and private as possible, but until we have a truly trustless E2E encrypted execution environment I don't see how compute offload technologies gets around this problem.
[0] > And to bring this model to production, we work with both Google and Nvidia to extend our Private Cloud Compute infrastructure to NVIDIA GPUs in Google’s cloud, while maintaining Apple’s unmatched privacy guarantees
https://9to5mac.com/2026/06/08/craig-federighi-details-apple...
[1] https://www.rfc-editor.org/rfc/rfc9458
2 replies →
https://en.wikipedia.org/wiki/PRISM
Exactly. That page objectively proves that Apple can't be trusted and breaks their promises.
https://security.apple.com/blog/expanding-pcc/
I've been a paid subscriber to Claude for a couple of years, but lately I've been reaching for the free Gemini app on my Android Pixel 9 because it's so good at doing searches as part of its answers. The model feels fresh and up to date. Whether Apple can incorporate that search is an open question
A lot of people are missing that Google is light years ahead in terms of edge AI. They've been going on about it even before the GPT-craze. Pixel phones have had live captions (on edge transcriber) for a while.
Apple ships iPhones with their neural engine since 2017.
Yet they’re lacking in the model spaces.
Things changed since Apple Intelligence but I was hoping there’ll be more things like live captions and what-not than chatbot use cases. I feel pixel is also moving towards that and abandoning the old way unfortunately.
How good is their offline (on-device) ai offering?
Very good. Just look at Gemma 4.
At its core, it’s still doing what Google Assistant and Siri were doing since many years
Not sure what extra are we achieving here
Only if you consider Google Image Search and Google Nano Banana to be "the same thing" since they both produce an image based on text input!
Similarly, Google Translate's millions of lines of hand-rolled code has been entirely superseded by LLMs that do a vastly better job.
The LLM-based AI assistants are based on a wildly different technology stack with very different capabilities compared to the legacy "if-then-else" logic programming that Siri was based on.
Was Google Translate millions of lines of hand-rolled code? The Transformer architecture was invented for Google Translate, before it was used to build "LLMs".
1 reply →
Better surveillance.
Maybe I'm wrong, but this seems to strongly undercut Apple's claims about privacy.
>The company reiterated that Apple Intelligence relies on on-device processing and Private Cloud Compute, with a promise that user data is only used to execute the immediate request and is not accessible to Apple or third parties. Apple added that outside experts can verify those privacy guarantees "at any time."
If you want to independently audit it from the outside, then you might not be an expert for Apple
> Apple added that outside experts can verify those privacy guarantees "at any time."
To my understanding, these "outside experts" have to go through a vetting process by Apple first. There are no publicly available audits of the infrastructure planned right now.
I don't think so. They will be running on their servers, or running in the future on Google's servers with privacy guarantees.
Do you think Google doesn't protect privacy for large paying customers?
For years I have enjoyed using Google products that I pay for, and they are clear about privacy guarantees.
This move kind of reminds me of the original iPhone with google maps. You're competing with google, but you're using their infrastructure. Why wouldn't they just go with another provider like OpenAI or Anthropic?
To be fair, with the original iPhone they weren't competing. Google did backend web stuff well, so Apple partnered with them and made nice local apps that were fed by Google's cloud data. The same was true for the YouTube app.
After Google saw the iPhone (before the public), they pivoted their vision for Android (it was originally on blackberry-style hardware), and that's when the "thermonuclear war" started. Kind of interesting Steve Jobs would have showed the iPhone off like that, when something very similar happened with Bill Gates, which prompted the development of Windows (according to Jobs).
19 years later, it's probably time to be pragmatic again. If Apple isn't able to deliver on some of these AI integrations into the OS, they risk losing users to Android. If they have to pay someone for a model, they might as well choose the one they think is going to be best for their users. This keeps existing iPhone users on iPhone, and may pull over some Android users looking for the same features, but with better privacy. That seems like a win for Apple. To pay OpenAI instead of Google would just be spite at this point. Maybe well deserved, but the leadership has all changed over in the last two decades, so they'd be fighting old wars. Though I think they should still think twice before showing Google anything that hasn't yet been released.
> 19 years later, it's probably time to be pragmatic again
These 2 have always been very pragmatic. They compete in an area or two but there is also a lot of money flowing both ways.
Given that they had originally selected OpenAI for Siri, and that deal fell through, I would guess something about their relationship with OpenAI fell through. Maybe OpenAI wouldn't let Apple run their model on Apple's servers.
OpenAI or Anthropic are not anywhere as well funded as Google. Apple already has everyone in their pocket via the ecosystem, they just have to not crap the bed. They value stability over the competitive component here.
Google is almost closer to a conglomerate than a coherent horizontally integrated corporation. The individual parts of Google are like Fortune500 companies themselves, and tend to act in their own interest.
OpenAI and Anthropic, despite bombastic media coverage, are still frontier labs. Google won't go down under if Gemini doesn't sell enough anymore. Apple and Google are planning for the aftermath of the AI IPO craze, one way or another.
Pretty much all of big tech compete with each other in some areas and partner in others.
Google probably gave them the best deal. When you're the #3 player you'll sacrifice margins to drive volume.
not just google maps but google search itself has been on the iphone since it launched
Are they competing with Google?
What do you consider Android to be, if not a competitor of iOS?
3 replies →
yeah or Magento teaming up with the X-Men to defeat the military in X2. XD
I’m not thrilled about any sort of Apple AI. I see it more as the convenience of platform lock in that ideally would be in the hands of all serious AI contenders but we all know that’ll never happen.
Every mainstream product seems to have their own “SmarterChild on steroids” bolted on top (Gemini for Google, Rovo for jira, Copilot for Microsoft everything, etc).
I’ll still use the serious ones like ChatGPT/Claude as my main but I think these companies know that and are just trying to jump the bandwagon so they don’t look outdate. Either way, they can be surprisingly convenient and make up for UI/UX learning curves.
I sniff iPhone prompts ending up in Musk's xAI servers...
Apple -> Google -> xAI datacenter
see: Google to pay SpaceX $920 million a month for compute capacity at xAI data centers
https://www.cnbc.com/2026/06/05/google-to-pay-spacex-920-mil...
They're using Google compute, that doesn't mean Google can access the data, especially given the E2EE setup Apple are going with.
> especially given the E2EE setup Apple are going with.
I'm endlessly amused by so-called hackers that insist "E2EE is secure" but forgot about "SSL added and removed here :)"
You really trust Google to protect your privacy in 2026? Because E2EE keeps you "safe"? Yeah, good luck with that: https://upload.wikimedia.org/wikipedia/commons/f/f2/NSA_Musc...
I used to wonder what "apps" might become in an "App Intent-first" world.
Bundles that provide data and capabilities to iOS and Siri? And perhaps libraries of UI components to display and interact with said data?
But then, if that works really well, and gets strong adoption, why ever open the app? What’s the point of having navigation flows inside an app? Could one make entire apps solely dedicated to providing a set of data, capabilities, and UI components to the system?
In that world, what drives user retention, for such apps? What even is an app? App engagement disappears as well.
And that’s not even diving into the use-case of Siri, say, planning a trip across five different apps (flights, hotel, restaurants, whatever) using just App Intents. If done well.
In that world, do most apps just become plugins, providers for Siri?
If you remember Windows Phone 8, it has already been tried and nobody (ahem, no company) wanted that.
No company is stupid enough to give up their content and infra and get none of the screen real estate.
I can see a parallel with hotels and OTAs, but in that case appearing on an OTA brings in sales. Showing $userA's content on $userB's screen won't earn any money from $company.
I remember Windows Phone 8 existed, but that’s pretty much it. And yes, that’s the big question: what’s in it for the app publishers?
:/
Well, there’ll always be room for people who want different UI/UX. Humans are too visual to ever move to “pure voice” and so we’ll inevitably have nice screens and thus UI preferences (technically you get preferences with voice too but it’s weirder)
Yes, but you can have this by providing surfaces as UI components to Siri.
Not sure if they do that (yet), but no reason an app couldn’t expose "Here’s what you can use to present data of shape X", or "here’s a UI for process doing y".
It feels like turning the common approach inside-out. But it works.
Edit: you could even imagine, in that world, apps that only expose surfaces, composable UI libraries, multi-step flows, declaring what they’re for, what kind of inputs they take, and what output they produce. Without ever owning any of the data (eg flights data, hotels inventory, booked trips, financial data, etc) or capabilities (eg book a flight).
The whole idea of an app becomes a much more fluid and transient thing
Everyone can have their own “view” into the data
I'm happy with this so long as the cloud side of things can be entirely disabled.
That is very logical. Google's AI has rapidly become better at search and presents answers in a tolerable way.
If we disregard exploitation of the primary sources and focus purely on a use case, search summaries are it. Generative AI is completely useless and OpenAI and Anthropic will soon fail. The Segway is maybe fun for a day, then you ask yourself why the heck you aren't using a bike in the first place.
I must use Google's AI responses on its web search about 40% of the time now.
It's just far easier to get a direct answer to what I'm looking for than going through mountains of SEO spam.
I congratulate SEO optimization companies on doing the very opposite of what they intended to do.
Everyone assumes this is apple paying Google to use their stuff. But maybe its the other way around. Maybe this is the google search engine deal 2.0. i.e. Apple agrees to sell their customers to Google for money. I know that right now its supposed to run privately and all that, but its still Google's voice telling you things, not to mention intepreting your instructions.
Lmao there is no way. Think about the economics of AI inference vs. Google search.
The latter was cheap and easily offset by ad views. The former is expensive, not to mention Apple has no internal prowess in this area atm.
And honestly? That's fine. One company shouldn't be "doing everything" no matter how desperately Apple wants that to be the case for itself.
Google right now offers plenty of AI inference for free for everyone on the planet. They also offered their £20 AI pro package for free for over a year for all students. I use that and its a very generous offering. Some more free AI inference for Apple users (where you've now selected for high income folk, simply by virtue of being Apple users) would make plenty of sense, I imagine.
1 reply →
not much info to personally go with tbh.
for those commenting on relying on a competitor for a "major" component of their product, look no further to how they source their displays from samsung all this time. even their silicon at one point! [1]
while it is now all about harness and "holding it right", their approach seem no different from other instances on hardware side. in terms of why not a more "leading" provider, look how the openai integration panned out so far. moreover, they have little to no initiative nor incentive to provide edge models or a hybrid solution so far. and anthropic's moat is not in audio.
[1] https://news.ycombinator.com/item?id=6419506
Any word on pricing for Private Cloud model usage? (It's only free if your app has had <2 million downloads, and it has rate limits.)
Answer: once you hit 2 million downloads on any app, you fuck off
Runs on Apple's private cloud compute, but is not available to Apple... That seems like a contradiction.
It does, I agree. That said, they've published some Apple Research docs/papers & core documentation where they outline the architecture and how it works. Personally I think their approach is fascinating.
https://security.apple.com/documentation/private-cloud-compu... https://security.apple.com/blog/expanding-pcc/ https://security.apple.com/blog/private-cloud-compute/
This must be a huge win for Google's cloud business. Getting Apple as a customer for Gemini is massive.
It is weird and disturbing that Apple has no native AI capability.
This is one of the most cash rich companies in the world and it has failed to have any position in the most critical technology development perhaps ever.
It's a clear signal that Apple became the most incredible operational/execution company under Tim Cook, but lost its innovation leadership.
You don't stay the most cash-rich company by chasing every expensive fad and they've been equally conservative with other "THIS WILL CHANGE EVERYTHING" tech fads such as Cryptocurrency and VR. I don't blame them for not rushing to light Billions a month on fire like the other big players; their play always seems to be to let things shake out and then deliver something refined and sophisticated.
There also doesn't seem like any real opportunity for them to Apple-ify this tech (any more than today's announcement). There's lots of rough edges and the underlying technology is fundamentally janky and extremely problematic in Apple's second differentiator of privacy.
They did light billions on fire for VR, just not as publicly
1 reply →
I take it as a signal they don't see any of their value being provided by the models. They're strong point was never frontier technologies. It's always been the delivery of the technology.
They are a hardware company at heart, the do software as software is needed.
For building a competitive AI they'd have to hire the talent, which is expensive and then do a massive investment, which may still end up far behind the competition. (See there attempts with Siri)
Now they can pick the model they want and if time is right they can still build their own.
In the end they still want to sell devices. They aren't doing a search engine (while they could), they are not doing an LLM model, ...
>> They are a hardware company at heart, the do software as software is needed.
This has never been true, not since Steve Jobs returned.
The heart of Apple is software and hardware integration.
Hardware companies that do software just to prop up the hardware business do terrible software, and (no doubt the Apple haters gonna hate this) but Apple does - for the most part - amazing software.
>> which is expensive and then do a massive investment
Apple has $145 billion in cash.
2 replies →
Come back when most of AI model makers go belly up....
Apple largest acquisition in their entire history is three billion dollars (Beats), Sam OpenAI got nothing and Google is getting a measly one billion dollar refund for Gemini.
Innovation is Apple Silicon and the five ecosystems (Microsoft and Nvidia aren't sparking now for nothing), Innovation is being the last American vertical computer company left from the 1980's who has been able switch gears went necessary (the next gear shift will probably be memory)...
Well it's half-and-half, why did Apple struggle for so long with Siri and its pre-LLM era technology, during the time of AlphaGo and so forth, and then after Covid why didn't Apple pivot to something like their own version of Gemini?
But there are lots of differing possible reasons for this, and I think it is premature to conclude with any one in particular.
I have been using their on define AFM models for a year - for small models they are good. Their Secure Enclave server bases AFM model is good, but not in the same class as gemini 3.5 flash or deep seek v4 flash.
Really I don’t think this is a strong take at all. If anything this has positioned them extremely, extremely well for when the bubble bursts and they can go with the winner to provide reasonable capabilities.
It's looking like an incredible move so far. They will have wasted no money while the whole industry lit trillions on fire, and then at the end they can just rent a model for cheap and replace it with their own or whatever is cheapest at any time.
The users really aren't Gemini users, they don't care what the model is behind the scenes.
They should have built a separate phone around Apple Intelligence. That is a no brainer. Jobs would have done the same. That is not going to be easy, who knows about user experiences better than Apple does (or atleast used to). This is underwhelming from Apple.
Remove all the apps and icons. All apps must just become skills in the background. Everything operated through voice or chat.
If Apple could crack chat intent to app action with 100% success rate. They have a product line unmatched and uncontested, for a decade.
Maybe Jony Ive & Sam Altman will crack it.
Privacy angle is the real story here. How much user data gets shipped to Google's servers vs kept on device? That's what I wanna know.
Wait, if it's Gemini why do they call it "Apple Intelligence"? Is Google okay with that?
Google sold Apple the ability to run certain Gemini models on Apple's data-center hardware, using Google's orchestration layer. Apple hooks into that not dissimilar from an API-provider, and then builds everything upstream.
Meaning the system prompt(s), harness, entry and exit points, and skills. So the product is still "Siri AI", because of all the stuff that takes it from a raw infrastructure concern upon up into a "product" is Apple's responsibility.
Google are "okay with that" because Apple pays them $1B a year, per press reports, to be.
Any business can do this now actually - you just need to lease/rent hardware through a Google partner and you can run an Nvidia based server in your own datacenter running (supposedly) the latest full Gemini models.
2 replies →
I don't think you conceptualize Google's game plan. all these companies care about is b2b contracts so they can inflate their balance sheets because when it's digital, it doesn't have to actually exist for it to "make money"
If iCloud is implemented on Google Cloud Storage, why do they call it iCloud?
I'm sure they're getting a sufficiently big paycheque to white label their models for the biggest consumer computer seller on the planet.
There's loads of AI features out there that are powered by a model provider, yet are not branded by them. Why would this be different?
People keep saying Gemini but it's not clear that the models are Gemini. They might be separate models.
Let us hope the EU forces Apple to allow the end-user to choose the external model: Wouldn’t it be amazing having privacy first local models calling out via a welldefined open protocol to a model of your choice: Claude, Grok, DeepSeek?
Sounds like OS architecture done right - screw the kickback business model.
That's basically what we built at Tinfoil. We run open source models inside secure enclaves (also using Intel TDX/AMD SEV-SNP + NVIDIA Confidential Computing). All the code running inside the enclave is open source and the client SDKs (also open source) automatically verify that the pinned source code matches the runtime attestation. The protocol used is TLS (terminates in the enclave) + HPKE keys generated inside the enclave on boot. Docs walk you through the verification process: https://docs.tinfoil.sh/verification/verification-in-tinfoil
Of course, we can't support Claude or Grok as they are closed source, but there is no incentive for companies that need your data to train the next generation of models to allow for private inference. One day...
Apple isn't shipping this in the EU
Funny, OP wants them to force a feature via regulation, and regulation is the reason they won’t even deliver the feature in question. Death by regulation.
3 replies →
Apple isn’t giving up the market of neither the EU nor China.
3 replies →
Let's actually not hope that, and let's not indulge the EU regulators' fantasies that they get to dictate the product design of products from non-EU countries.
Force Apple to support all kinds of arbitrary models? That's a comically bad idea.
“Let’s not indulge US regulators fantasies that they get to dictate the product design of cars from non-US countries.”
Of course governments have the ability to decide what products are sold in their countries and how.
2 replies →
Why would it be ok that a monopoly business prevents free competition and consumer choice by only allowing certain or a single model provider that likely gives them kickback via some opaque business deal?
> EU regulators' fantasies that they get to dictate the product design of products from non-EU countries.
They should get to regulate the design of products from non-EU countries only those that are sold inside EU.
The fact that it is not cost effective for Apple to design two separate products(software or hardware) for EU and non-EU is an Apple problem.
Oh nice, maybe they'll finally make Siri learn foreign languages?
they can hardly imagine a world outside SV, let alone the US. I'm not holding my breath
I tried the on-device image input model with an app I am building. It's not very good at world-knowledge recall, but can describe the submitted image well enough.
Sort of expected this at the first attempt. Use their existing partnership for Google being the default search with Google and leverage Google's models.
Sorry to be off topic, but I have a question: has anyone installed the latest beta iOS and macOS, and if so what is the current status of Gemini integration?
once you update to the latest iOS, there's an option to join the waitlist. No indication of timing, but until that moves forward Siri is what you get today, as far as I can tell.
Built around=Front end?
Google apps are the most downloaded apps in the Apple App Store already. This reminds me of the original Apple Maps, which was just a front end for Google Maps.
I'm wondering who's paying for this. Is it becoming a part of the iCloud subscription? A separate billed product?
I'm really not looking forward to Gemini models on my devices.
Gemini models clearly gaslight the user and hallucinate, they're also SUPER verbose, as shown in the demos from the keynote.
Plus, if they're not charging a subscription for this, you know we're getting the dumbest models...
If you pay for Gemini, then it is good. I recently used Gemini Ultra for a month and the gemini models are very good (and of course, you get a lot of Claude Opus tokens to use through the same plan).
I also pay for Proton's Lumo+ private chat and for what it is it is also good.
The free plans from all the providers are bad, which is fare enough.
I use Apple devices and I expect to be paying for Gemini tokens after the integration.
It's a little interesting our duopoly of mobile phone OS controllers are so closely integrated with each other.
Just as long as you speak a major language
I expected a beter infrastructure for this new AI. I hope they build it to be more robust slowly
The full Apple prompt has leaked online. It's on github and reddit. From ios logs.
Will it be possible to disable, or better yet, wipe the whole feature off?
Last mover advantage.
Just an interesting thought on the very different philosophical approaches. Let's imagine that Android had a terrible assistant, so bad that even most Google Fanbois admitted it was pretty bad. Apple became a leader in AI. Google approaches them to license the Siri model for Android. Would Apple have ever done that?
So they run on TPUs and not Nvidia chips?
The heaviest models run on Nvidia
No, Siri runs on Apple Silicon.
“Now, we are collaborating with Google and NVIDIA to run new Apple Intelligence workloads on Google Cloud, extending our industry-leading PCC privacy commitments to third-party data centers for the first time.”
https://security.apple.com/blog/expanding-pcc/
https://x.com/nvidia/status/2064100362752294992
Thanks. That's still very interesting that they don't need Nvidia to do any of this. Nvidia stock has been priced like AI isn't possible without them.
2 replies →
Looks like AI reached it's peak already. Every new architecture is just a new wrapper that has nothing to do with ml.
Not launching in EU feels like a smell. It does look interesting enough for me to try it out before disabling Apple Intelligence again.
It’s the DMA regulation that forces Apple to give the same access as they have to other AI chat apps.
Once it leaves the device Apple does not know what those other ai chat apps will do with the gathered data.
> Siri AI is private by design and deeply integrated across Apple’s platforms using on-device processing and Private Cloud Compute, which extends the privacy and security of iPhone into the cloud. However, under EU regulators’ extreme interpretation of the DMA, Apple would have to give any virtual assistant direct access to users’ private data — and the ability to directly control other installed applications — as soon as Siri AI is made available in the EU, without the essential protections necessary to keep users and their data safe.
https://www.apple.com/newsroom/2026/06/due-to-dma-siri-ai-de...
Apple loves to play dumb about this stuff. The EU imposes a pretty straightforward regulation regarding equality of access. Apple seems to come up with all sorts of "solutions" to this "problem", and each one never amounts to true equality of access. They could easily just allow users to decide "Do you want to give this app unfettered access to all your device data, including other apps' data?". Let users decide. 99% of Apple users in the EU will probably click "no". I'm sure they'll make the user warnings scary enough to ward off anybody who doesn't know what's going on.
There are 2 potential outcomes: either the sky really does fall, and there's a meaningful uptick in bad things happening to iPhone users, in which Apple can easily point the finger at the EC and say "they made us do this". Apple looks like the good guys who put up a good fight for their users, but ultimately their hands were tied, and they'll probably get the revisions to EU law they're so desperately fighting for.
The other possibility is that the sky does not fall, and Apple looks both silly and malicious at the same time for ever having suggested that it would, which was clearly in bad faith.
Clearly, Apple cannot afford scenario #2, so I think they will probably never give their users the actual freedom that the MDA requires them to. They will just exit Europe entirely before allowing that to happen.
15 replies →
> EU regulators’ extreme interpretation of the DMA
It's not extreme interpretation, it's the intent.
Just say it would break your vendor lock-in.
10 replies →
so to translate:
- Apple has powerful capabilities in iOS to enable Siri AI.
- EU's DMA requires them to allow users to install third-party AI backends.
- Apple doesn't think parties other than themselves should be trusted with those iOS permissions.
I guess it'd be like if Apple allowed a first-party screen reader for iOS, so they refused to allow third-party screen readers.
15 replies →
> Once it leaves the device Apple does not know what those other ai chat apps will do with the gathered data.
It's the user's data. Not Apple's. And it should be the user's right to send it to whoever for whatever results, imo
11 replies →
> Once it leaves the device Apple does not know what those other ai chat apps will do with the gathered data.
Yeah, that's the whole fucking point.
> It’s the DMA regulation that forces Apple to give the same access as they have to other AI chat apps.
But why can Tesla ship Grok to their cars in the EU without any problems? Why aren't they required to let me choose between Grok, OpenAI etc or even a custom endpoint?
5 replies →
Translation:
Since it's the user's device, not Apple's, EU correctly "interprets" this as the user has the right to do whatever they please, including installing third-party chat apps.
Apple are just bulshitters when it comes to actual users, and not their corporate definition of a user.
BTW, did you know that in Japan, and in Japan only, you can change the Siri shortcut button to start other voice assistants? https://mjtsai.com/blog/2025/11/18/ios-26-2-third-party-voic...
Or that they wouldn't let you set default maps app outside of the EU: https://mjtsai.com/blog/2025/03/14/dma-compliance-default-ma...
12 replies →
Sounds like Apple PR bullshit.
Unless Apple proves otherwise I'm more inclined to believe they're either 1. Using this to try and shape the DMA in their own interest (definitely not their users' interest) or 2. Doing something with the data that would not be allowed in the EU (also not in their users' interest at all) or both.
Because outsourcing to Google is so much trustworthy...
14 replies →
They say it's because the EU's DMA would require them to open up device data to third-party assistants, and they'd no longer be able to guarantee user privacy.
https://www.apple.com/newsroom/2026/06/due-to-dma-siri-ai-de...
I don't see what the issue is. The user could then select Apple (or Mistral) for strong privacy or another provider for customers that don't care.
I primarily want Apple to provide extension points so that I can select my own provider, just as I can choose where to host my mail or install another app as my instant messenger.
Sure, I could install another provider's app, but it wouldn't have the same integrations, similar to how an instant messaging app would be less useful if notifications were limited to iMessage.
1 reply →
I'd love to select Mistral over Google Gemini.
2 replies →
Sure.
It is a smell. But it's the EU that smells bad, when it comes to tech regulation. It's the smell of cookie popup warnings.
Nothing in the law requires the pop up. It definitely doesn’t require the obnoxious bullshit that most companies put up (aka the dark pattern to get you to agree to every unreasonable part of their terms just to read the page).
The alternative would be to just stop invasive tracking and add the cookie when it’s actually needed.
6 replies →
The smell is that Apple doesn't want to give the same level of access to third-party AI assistants that Siri will get.
For what it's worth, Apple claimed they proposed an "equivalent access" framework with some kind of "trusted agent framework" approach, but that it was shot down by the EU. I suspect it was way more inconvenient for third-party developers than Apple lets on.
This has been the case for quite a while. Like Reminders - You can't replace the phrase "Siri, remind me to ___" with a third party app. I'm surprised the EU lets them ship Reminders there.
2 replies →
Nothing obligates companies to face steep EU regulation and fines and launch there from the get go.
If anything it should concern fellow Europeans that consumers are paying more for less and later.
[dead]
what i think why they are too much relying on Google is coz they are way much towards making models open source and launching more much better models to public as if in future apple part way from google they might still have much better models to rely on and if we see the history google has been partnered with Apple since the launch of first Iphone
Another one bites the dust
What a blunder, they resisted AI for like 2 years when it was all the buzz, and now when the bubble is about to bust and every user has AI fatigue they decide to finally dip into the fad?
Before it was as if avoiding AI was a conscious design decision, and if there was an AI crash, Apple would be the only survivor left. Now it feels like they weren't in on the meme out of incompetence and are now late to the party.
No one can know what Jobs' stance would be, but I like to think he would be anti-slop
I agree that many AI businesses will go bust and they deserve it, but the tech is good.
I can recommend my own layered approach, using the lowest capability models that get stuff done:
1. I maximally use local models like gemma4:26b-a4b-it-qat for everything that works with this free option.
2. I like paying for inexpensive APIs for mid-tier models like deepseek v4 flash, gcp-5-mini, gemini-2-flash for things that option 1. fails at. This option is almost free.
3. Pay for more expensive APIs like deepseek v4 pro, gemini 3.5 flash, etc. This option is not too expensive.
4. If all else fails on a class of tasks, then pay for awesomeness of Claude Opus. $$ expensive, I try not to use unless absolutely necessary.
I think developers and companies that just cram everything into Claude Opus are unprofessional.
I think you are conflating LLMs with AI.
LLMs we all agreed were amazing back in 2023-2024.
What's happening now with AI is more of a corporate phenomenon quite removed from the actual tech.
Yes LLMs are useful, but replacing customer support with an LLM that gives user accounts away, or calling LLMs on a loop where the bottleneck is your checkbook and calling it AGI, those are phenomenons that are separate from LLMs.
1 reply →
When people say the AI bubble is about to bust, I don't think anybody means that "the use of AI is going to go away." AI is absurdly useful. I think what people mean is "the valuations of these companies will have to snap to a reality that is actually attached to their market value."
Exactly, small edge models is the future, highly personal experiences, and not these massive models that the cloud providers currently shove down our throats. While massive models are useful, these massive platforms are about to burst out of their promises. All while we’re supper happy with tiny 4b up to 12b models working amazing for all these “omg ai thinks” daily tasks.
Go Genmoji!
Spy phone
Guggupple
ok i will try thanks
Maybe now we'll get a good voice prompt experience with Gemini on iPhone out of this deal.
I honestly don’t understand how anyone can believe that Apple is limiting user options for privacy reasons, rather than trying to maintain an unfair advantage over other vendors.
I’m not saying people who hold this view are being dishonest at all. But sometimes, to me, brands like Harley-Davidson or Apple seem closer to a cult than to a typical corporation.
Because privacy is actually a big feature, many people are skeptical about AI and the big model providers and don't trust them. They're less skeptical about on-device AI and so Apple is pushing that and making privacy a core feature of their online offering as well.
I probably wouldn't use it without that. It's one thing sending my shitty code to be trained on, and another thing entirely to give these companies access to my personal life and information.
That still doesn't explain why my data can only be sent to Apple and not to another vendor of my choice.
8 replies →
> ...closer to a cult...
When you have to image a highly irrational reason to explain why groups of people do the things they do, there's a decent chance you just don't understand their perspective. They may be acting reasonably rationally from their own perspective. (As you said yourself: "I honestly don't understand...")
One can be rational but still be in a bubble or cult.
2 replies →
Uh oh What is Google getting from this?
There was a meme the other day that said: the US economy right now is 7 companies giving each other trillions. I think this that that pattern.
[flagged]
[flagged]
[dead]
[dead]
[dead]
[flagged]
Nah, they crossed the line when they allowed photo editing apps on the AppStore which can do all the things you listed and more. It’s disgusting.
[flagged]
6 replies →
Apple could have done something like bedrock and used a SOTA model but instead they are fiddling with local models or whatever.
Also I have seen that Apple has some strange lust towards image generation as if that's what people really want. I have this slop image generation thing on my phone and it is useless.
Here's what I want: natural language interaction to achieve complex workflows in iPhone. Example: find the cheapest way to go from A to B and book it using the Deutsche Bahn Train app.
Local Gemini is a light year ahead of whatever Siri is so this is a big improvement already.
If they don’t like this in the future they can just change to the less convenient, less secure, and likely more expensive bedrock + SOTA.
I don't disagree that local Gemini is better but if you've tried it in iPhone, it is slow, hallucinates and overheats the phone. For anything slightly non-trivial like the workflow example I gave, I think it will be close to useless.
[flagged]
This is coming very close to a personal attack, which is against the site rules.
Even if it's a bad take, call out what's wrong with the take, rather than attack the author.