Wasn’t it the same with covid hiring? While others over hired, Apple was modest in this position. Then everyone needed to significantly downsize, when Apple didn’t.
From a financial market perspective, AAPL is the second highest valuation for a publicly traded company and #1 is in first place because of the AI bubble.
Looking at how others stuff AI into everything they can, user experience be damned, I’m kind of glad Apple was perfunctory in its jump on the bandwagon.
Let everyone else pay for the research and make the mistakes, find out what works and what doesn't. Apple already has the consumers, they might as well save a few (hundred?) bn in the process and later deploy something which doesn't tell you to glue your cheese to your pizza.
Or like with the M1 chip: wait until the incumbent alienates so many experts in the field that you can scoop them up and they will succeed partially fueled on spite against their old employer.
They are flat-out incompetent. Siri has somehow regressed over the years and visual intelligence only works in demos. They have the most abominable integration with ChatGPT imaginable.
At least the MLX team has been shipping an impressive product.
I genuinely never understood why there was a narrative that Apple is "falling behind" when it comes to AI. They make phones, computers and an ecosystem of services to lock you in. None of this stuff is threatened by AI; with the right integration, it would enhance them!
Or, since the stock market is an emotional game (hear me out): Apple hasn't announced anything in the past year which caused comparable excitement and resulted in (further) overvaluation of their company like it happened on Microsoft, nVidia, etc.
Now, after a few months (!), reality sets in and those hyped-up investors realize that it's not as much of a short-term game as they told themselves it would be...
Apple has a p/e of 38, Nvidia is 46, Microsoft is 34. S&P has historically averaged around 20, so on that metric Apple and Nvidia are more similar than different.
Ah, the benefits of having a sober old CEO. And a business model that doesn't need to be buoyed nor stabilized by spinning and hyping a succession of hot new trends.
Apple doesn’t own a search engine either, and gets $20B per year from Google to direct search queries to them.
I hope they adopt the same model with AI - leverage whatever frontier model is best and provide their own privacy infrastructure in front.
At some point Apple will figure out a way to provide the right info from your calendar, messages, email etc as context and couple this with a bunch of secure tools for creating calendar entries, etc. Agentic AI will then be something I personally benefit from.
This is the thing I've found amazing about people's complaints about Apple and AI.
Historically the strength of Apple was that they didn't ship things until they actually worked. Meaning that the technology was there and ready to make an experience that was truly excellent.
People have been complaining for years that Apple isn't shipping fast enough in this area. But if anything I think that they have been shipping (or trying to ship) too fast. There are a lot of scenarios that AI is actually great at but the ones that move the needle for Apple just aren't there yet in terms of quality.
The stuff that is at a scale that it matters to them are integrations that just magically do what you want with iMessage/calendars/photos/etc. There are potentially interesting scenarios there but the fact is that any time you touch my intimate personal (and work) data and do something meaningful I want it to work pretty much all the time. And current models aren't really there yet in my view. There are lots of scenarios that do work incredibly well right now (coding most obviously). But I don't think the Apple mainline ones do yet.
What people hate about Apple is that they ship things other people couldn't get to capital-W Work, and they're seen as 'stealing' the idea instead of perfecting them.
Historically the strength of Apple was that they didn't ship things until they actually worked. Meaning that the technology was there and ready to make an experience that was truly excellent.
It might as well be the visualization of the two strategies:
- Everyone else: "We mainly build huge AI compute clusters to process large amount of data and create value, at high cost for ramp-up and operation."
- Apple: "We mainly build small closed-down AI compute-chips we can control, sell them for-profit to individual consumers and then orchestrate data-processing on those chips, with setup and operational cost all paid by the consumer."
I can't think of any company which has comparable know-how and, most of all, a comparable sell-out scale to even consider Apple's strategy.
No matter what they do, they will sell hundreds of millions compute devices for the foreseeable future. They use this to build out AI infrastructure they control, pre-paid by the future consumers.
> I can't think of any company which has comparable know-how and, most of all, a comparable sell-out scale to even consider Apple's strategy.
I'm not sure where you position Samsung or Xiaomi, Oppo etc. They're competitive on price with chipsets that can handle AI loads in the same ballpark, as attested by Google's features running on them.
They're not vertically integrated and don't have the same business structure, but does it matter regarding on-device AI ?
I recently tried to figure out what their offerings currently are. I'm hoping for `efficent but performant AI compute-chips` by Apple ever since they kicked out Nvidia in 2015 (for the ML Models / Exploration parts bellow). It will be interesting to see how good their products will feel in this fast-paced environment and how much legroom (RAM + Compute) will be left non-platform offerings.
To my understanding, they market their ML stack as four layers [1]:
- Platform Intelligence: ready-made OS features (e.g., Writing Tools, Genmoji, Image Playground) that apps can adopt with minimal customization.
- ML-powered APIs: higher-level frameworks for common tasks—on-device Foundation Models (LLM), plus Vision, Natural Language, Translation, Sound Analysis, and Speech; with optional customization via Create ML.
- ML Models (Core ML): ship your own models on-device in Core ML format; convert/optimize from PyTorch/TF via coremltools, and run efficiently across CPU/GPU/Neural Engine (optionally paired with Metal/Accelerate for more control).
- Exploration/Training: Metal-backed PyTorch/JAX for experimentation, plus Apple’s MLX for training/fine-tuning on Apple Silicon using unified memory, with multi-language bindings and models commonly sourced from Hugging Face.
> They use this to build out AI infrastructure they control, pre-paid by the future consumers.
I'm not following. What infrastructure? Pre-paid how?
Apple pays for materials and chips before it sells the finished product to consumers. Nothing is pre-paid.
And what infrastructure? The inference chips on iPhones aren't part of any Apple AI infrastructure. Apple's not using them as distributed computing for LLM training or anything, or for relaying web queries to a complete stranger's device -- nor would they.
Apple's phones are responsible for most of their revenue. The phones are designed to pretty much exclusively interact with social media and take photos. AI doesn't really add anything to that experience since advertisement consumption by humans is the ultimate objective. That's why even though Apple's Siri has been about the most useless assistant in existence for years, Apple isn't in a rush to replace it. It simply doesn't have a big impact on their revenue.
Microsoft has been criticized for investing in AI heavily. But it actually makes sense for Microsoft if you consider the nature of their business. The problem is not with the investment per se but with what they got out of it. Unfortunately, Microsoft sucks at product management, so instead of creating useful stuff that users want and are ready to pay for, they created stuff that no one understands, no one can use, and no one wants to pay for. Github copilot is an exception of course. I'm talking more about their Office 365 AI.
The revenue from AI is growing at a much slower rate than recurring capex and depreciation is accumulating. This will create distress opportunities that cash-rich companies like APPL may seize. Might be a private equity deal, might be in the public markets as some of the players dip hard after IPO.
As this plays out, APPL's silicon has unified memory, power consumption and native acceleration that gives it an edge running SLMs and possibly LLMs at scale. Wouldn't shock me to see APPL introduce a data-center solution.
The core of Apple's problem boils down to apathy towards their product quality. I just recently switched from using Siri to Google Gemini in my car. The experience is dramatically better.
And this is the case across the board.
My friend's Fitbit works way better than my Apple watch.
Third and final example is how bad Apple's native dictation engine is. I can run OpenAI Whisper models on my Mac and get dramatically better output.
As a long time Apple fan who's had everything since before the first iPhone, I feel this apathy towards product quality cannot be disguised as some strategic decision to fast follow with AI.
But for everything else, you literally just said, the handful of AI features are better on Google products... That seldom makes the product as a whole better.
I am going to defend Apple: their new built in system model in iOS26 and iPadOS26 is very decent, similar to the small Google Gemma models and the small Chinese models. For complex queries a free API call is transparently made to a secure computed environment on Apple’s servers that are documented to preserve privacy.
A problem is that even though it is super simple to write Swift / SwiftOS apps to use the system model, I don’t see much evidence that many developers are using the model in their apps.
FWIW, AI is not entirely locked down in the Apple ecosystem. Sure, they control it but they've already built the foundation of a major opportunity for developers.
There's an on device LLM that is packaged in iOS, iPadOS and macOS 26 (Tahoe) [1]. They even have a HIG on use of generative AI [2]
Something like half of all macs are running macOS 26 [3] already, so this could be the most widely distributed on-device LLM on the planet.
I think people are sleeping on this, partly because the model is seen as under powered. But I think we can presume it won't always be so.
I've just posted a Show HN of app for macOS 26 I created that uses Apple's local LLM to summarize conversations you've had with Claude Code and Codex. [3]
I've been somewhat surprised at the quality and reliability of Apple's built-in LLM and have only been limited by the logic I've built around it.
I think Apple's packaging of an LLM in its core operating systems is actually a fast move with AI and even has potential to act as an existential threat to Windows.
I can second this. I am nearing launch on an app that uses both the new SpeechAnalyzer and on device LLM and it has met or exceeded my expectations. A longer context would always be nice but then I remember its running on a phone.
Don’t a lot of Android devices come with Gemini Nano on the device?
Probably not as many out there as there are Apple devices because it is only the high end ones at the moment. I don’t think they are that far behind in numbers though.
I've tested almost every LLM which will work on a modern iPhone and Apples models are universally terrible in comparison to almost every open-weights model, they're so bad it's a joke amongst devs who work in this space.
The only thing it's useful is super basic tasks like sentiment classification, summarization (sort of), or stuff like, "Does this message contain toxic/bad language, answer yes or no only".
Seems that this is apples modus operandi since the app store, their last "thing" they've made really.
Hype about self driving cars -> apple chases it with apple car -> investors pleased they kept up with the joneses -> apple car is behind or not good enough or whatever -> quietly cancelled -> investors pleased they culled the deadweight.
You can replace apple car with vision pro or soon apple intelligence and it will play out the same formula. Luckily it allows investors to profit.
76 comments
[ 3.6 ms ] story [ 83.3 ms ] threadFrom a user perspective it may not be a strength: users / customers may expect certain functionality that works accurately and responsively.
Let everyone else pay for the research and make the mistakes, find out what works and what doesn't. Apple already has the consumers, they might as well save a few (hundred?) bn in the process and later deploy something which doesn't tell you to glue your cheese to your pizza.
lmao, even
They are flat-out incompetent. Siri has somehow regressed over the years and visual intelligence only works in demos. They have the most abominable integration with ChatGPT imaginable.
At least the MLX team has been shipping an impressive product.
Now, after a few months (!), reality sets in and those hyped-up investors realize that it's not as much of a short-term game as they told themselves it would be...
I hope they adopt the same model with AI - leverage whatever frontier model is best and provide their own privacy infrastructure in front.
At some point Apple will figure out a way to provide the right info from your calendar, messages, email etc as context and couple this with a bunch of secure tools for creating calendar entries, etc. Agentic AI will then be something I personally benefit from.
Historically the strength of Apple was that they didn't ship things until they actually worked. Meaning that the technology was there and ready to make an experience that was truly excellent.
People have been complaining for years that Apple isn't shipping fast enough in this area. But if anything I think that they have been shipping (or trying to ship) too fast. There are a lot of scenarios that AI is actually great at but the ones that move the needle for Apple just aren't there yet in terms of quality.
The stuff that is at a scale that it matters to them are integrations that just magically do what you want with iMessage/calendars/photos/etc. There are potentially interesting scenarios there but the fact is that any time you touch my intimate personal (and work) data and do something meaningful I want it to work pretty much all the time. And current models aren't really there yet in my view. There are lots of scenarios that do work incredibly well right now (coding most obviously). But I don't think the Apple mainline ones do yet.
Great artists steal.
- Everyone else: "We mainly build huge AI compute clusters to process large amount of data and create value, at high cost for ramp-up and operation."
- Apple: "We mainly build small closed-down AI compute-chips we can control, sell them for-profit to individual consumers and then orchestrate data-processing on those chips, with setup and operational cost all paid by the consumer."
I can't think of any company which has comparable know-how and, most of all, a comparable sell-out scale to even consider Apple's strategy.
No matter what they do, they will sell hundreds of millions compute devices for the foreseeable future. They use this to build out AI infrastructure they control, pre-paid by the future consumers.
THIS is their unique strength.
I'm not sure where you position Samsung or Xiaomi, Oppo etc. They're competitive on price with chipsets that can handle AI loads in the same ballpark, as attested by Google's features running on them.
They're not vertically integrated and don't have the same business structure, but does it matter regarding on-device AI ?
To my understanding, they market their ML stack as four layers [1]:
- Platform Intelligence: ready-made OS features (e.g., Writing Tools, Genmoji, Image Playground) that apps can adopt with minimal customization.
- ML-powered APIs: higher-level frameworks for common tasks—on-device Foundation Models (LLM), plus Vision, Natural Language, Translation, Sound Analysis, and Speech; with optional customization via Create ML.
- ML Models (Core ML): ship your own models on-device in Core ML format; convert/optimize from PyTorch/TF via coremltools, and run efficiently across CPU/GPU/Neural Engine (optionally paired with Metal/Accelerate for more control).
- Exploration/Training: Metal-backed PyTorch/JAX for experimentation, plus Apple’s MLX for training/fine-tuning on Apple Silicon using unified memory, with multi-language bindings and models commonly sourced from Hugging Face.
[1] https://developer.apple.com/videos/play/wwdc2025/360/
I'm not following. What infrastructure? Pre-paid how?
Apple pays for materials and chips before it sells the finished product to consumers. Nothing is pre-paid.
And what infrastructure? The inference chips on iPhones aren't part of any Apple AI infrastructure. Apple's not using them as distributed computing for LLM training or anything, or for relaying web queries to a complete stranger's device -- nor would they.
Microsoft has been criticized for investing in AI heavily. But it actually makes sense for Microsoft if you consider the nature of their business. The problem is not with the investment per se but with what they got out of it. Unfortunately, Microsoft sucks at product management, so instead of creating useful stuff that users want and are ready to pay for, they created stuff that no one understands, no one can use, and no one wants to pay for. Github copilot is an exception of course. I'm talking more about their Office 365 AI.
The revenue from AI is growing at a much slower rate than recurring capex and depreciation is accumulating. This will create distress opportunities that cash-rich companies like APPL may seize. Might be a private equity deal, might be in the public markets as some of the players dip hard after IPO.
As this plays out, APPL's silicon has unified memory, power consumption and native acceleration that gives it an edge running SLMs and possibly LLMs at scale. Wouldn't shock me to see APPL introduce a data-center solution.
Trying and failing to make a SoTA foundational model is not a strategic move. It's similar to Amazon and Meta, they also have tried and not succeeded.
And this is the case across the board.
My friend's Fitbit works way better than my Apple watch.
Third and final example is how bad Apple's native dictation engine is. I can run OpenAI Whisper models on my Mac and get dramatically better output.
As a long time Apple fan who's had everything since before the first iPhone, I feel this apathy towards product quality cannot be disguised as some strategic decision to fast follow with AI.
I went through three FitBits. After the third failed just outside warranty I got an Apple watch, which has outlasted all three FitBits.
But for everything else, you literally just said, the handful of AI features are better on Google products... That seldom makes the product as a whole better.
A problem is that even though it is super simple to write Swift / SwiftOS apps to use the system model, I don’t see much evidence that many developers are using the model in their apps.
There's an on device LLM that is packaged in iOS, iPadOS and macOS 26 (Tahoe) [1]. They even have a HIG on use of generative AI [2]
Something like half of all macs are running macOS 26 [3] already, so this could be the most widely distributed on-device LLM on the planet.
I think people are sleeping on this, partly because the model is seen as under powered. But I think we can presume it won't always be so.
I've just posted a Show HN of app for macOS 26 I created that uses Apple's local LLM to summarize conversations you've had with Claude Code and Codex. [3]
I've been somewhat surprised at the quality and reliability of Apple's built-in LLM and have only been limited by the logic I've built around it.
I think Apple's packaging of an LLM in its core operating systems is actually a fast move with AI and even has potential to act as an existential threat to Windows.
[1] https://developer.apple.com/videos/play/wwdc2025/286/
[2] https://developer.apple.com/design/human-interface-guideline...
[3] https://news.ycombinator.com/item?id=46209081
Probably not as many out there as there are Apple devices because it is only the high end ones at the moment. I don’t think they are that far behind in numbers though.
The only thing it's useful is super basic tasks like sentiment classification, summarization (sort of), or stuff like, "Does this message contain toxic/bad language, answer yes or no only".
Hype about self driving cars -> apple chases it with apple car -> investors pleased they kept up with the joneses -> apple car is behind or not good enough or whatever -> quietly cancelled -> investors pleased they culled the deadweight.
You can replace apple car with vision pro or soon apple intelligence and it will play out the same formula. Luckily it allows investors to profit.
That's not lucky. That's sad. They never ask the question "could we have earned _more_ profits with a better strategy?"
The market is not rational.