Office Hours #Q&A #WWDC #transcript
Recording and transcript of the June 9, 2026 Asymco One Office Hours, recorded the day after Apple's WWDC 2026 keynote. Horace Dediu shares his impressions of the event and takes questions from Asymco One subscribers.
Subscribe to Asymco One to listen to the recording and read the transcript.
Sign up and latest updates at asymco.com/live.
One Foundation Model runs on Nvidia GPUs hosted in Google’s cloud and some Google proprietary model data used to refine Apple's Models.
Apple has shared new details about the technology behind its latest Apple Intelligence platform, explaining how its new Apple Foundation Models (AFM) work and clarifying the company's relationship with Google. During a post-WWDC technical discussion, Apple executives said the company does not use Google’s Gemini models, Google Search, or Google Assistant infrastructure to power Siri or Apple Intelligence
Craig Federighi, Apple’s Senior Vice President of Software Engineering:
“Of course, we don’t have the Gemini app as our app. In fact, none of that client code is part of how we run on iOS. For these models, we use none of the models that Google deploy to their customers, nor do we use the infrastructure and means by which they deploy models to their customers. And then, when it comes to the knowledge base, we of course don’t use Google Search or anything like that as the foundation of our system. So I hope that’s clear. The amount of the Google Assistant we use is none.”
Source: The Mac Observer
NB: Apple introduced five foundation models.
These models are custom-built for Apple Silicon, trained using proprietary data, and refined using outputs from Gemini frontier models. The company emphasized that there is no direct use of Gemini itself.
Upgrades keep coming but are they justified?
Bank of America upped its price target on the tech giant, citing a potential boost from agentic artificial intelligence.
Shares rose 0.5% in premarket trading.
“In an agentic world, value accrues to the platform that controls user intent, personal context, app access, permissions, identity, authentication, payments, and trust,” analyst Wamsi Mohan wrote in a note to clients. “Smartphone is the scaled consumer device where these factors already converge. If AI assistants become the new front door to search, apps, commerce, scheduling, payments, and workflow completion, we think Apple should have meaningful leverage over model providers, app developers, merchants, advertisers, and payment networks.”
Mohan raised his price target to $380 from $330 and reiterated his Buy rating on Apple.
Delving deeper, Mohan said that the company's benefit in agentic AI comes from its semiconductors and iOS, as Apple's silicon can dictate what happens on a devices (which is important for latency, reliability, privacy and costs) and iOS will determine how the AI is used by the user. “Our analysis around token usage suggests Apple will need to rely on distributed compute (on device), Private Cloud Compute and third-party compute to deliver a robust AI experience,” Mohan added. “Removing the net cash neutral target, in our opinion, is a reflection of a new period of AI-driven investments to enable this agentic era.”As such, Mohan believes that the coming Siri redesign needs to be integrated into the iPhone and make it an agent that is capable of understanding intent, retrieving context, bringing up apps, and completing workflows. If this were to happen, an agentic version of Siri could boost fiscal 2030 revenue between $15B and $30B, or as much as $40B to $65B if users really became acclimated to it.
Source: Seeking Alpha
NB: Updated price targets with BofA change highlighted.
Horace speaks carefully and with knowledge about opinions, facts and fantasies, and shares what he thinks John Ternus’s most important job is.
Topics include:
Listen to Asymetric in your favorite podcast app.
Why give all your cash to Nvidia? Why not make your own chips? Then why not sell them?
Google said that it plans to sell its custom Tensor Processing Units (TPUs) to select customers who will install the chips in their own data centers.
The move is a change from Google’s prior strategy, which saw it rent out TPU capacity to customers from its own data centers — and is yet another strike at AI chip king Nvidia.
The announcement, during the company’s Q1 earnings call, comes a week after Alphabet announced two new TPUs: its TPU 8t for AI training and TPU 8i for inferencing.
Alphabet didn’t disclose potential customers, but it signed a multiple-gigawatt agreement for next-generation TPUs with Anthropic earlier this month, with chips expected to begin coming online in 2027.
And according to The Information, Alphabet has also entered into a multibillion-dollar chip deal with Meta (META).
Source: Yahoo Finance
NB: Alphabet Q1 sales grew 22% to $110 billion, with Cloud up 63% to $20 billion. Operating margins grew by 220 basis points, with services and cloud operating margins both increasing y/y.
The search business continues to benefit from AI. AI overviews and AI mode are sustaining as search sales grew 19%.
A lively discsussion of Apple topics such as:
Contact us at events@asysmco.com if you have any questions.
For most up-to-date information, see
Podcast #AI #Apple 50 #Disruption Theory #Nostalgia #Origin Story · 4 comments
Ben Thompson of the Stratechery interviews Horace on the Stratechery Plus Podcast to discuss Horace's journey to become an influential tech analyst: early career in research, business school under Clayton Christensen, and his time at Nokia during the rise of the iPhone.
The conversation further explores Disruption Theory in the context of both smartphones and AI, why Christensen misjudged Apple, and whether AI could challenge today’s tech giants. It also marks the launch of Dediu’s new paid service Asymco One and reflects his continued conviction in Apple Inc. at its 50-year milestone.
Oh the humanity
On Tuesday, Apple announced updates to its Apple Business program
Apple Business isn’t so much a new thing as a consolidation of what used to be two separate products–Apple Business Manager and Apple Business Essentials–into one. The new platform includes device management, company email and calendar at your own domain, an employee directory, and brand management tools across Maps, Mail, Wallet, and Siri.
The device management piece used to cost money and was only available in the U.S. While you still pay if you want to add iCloud storage or AppleCare+ for business, the core business tools are now free. “We’ve unified Apple’s strongest business offerings into one simple, secure platform,” said Susan Prescott, Apple’s vice president of Enterprise and Education Marketing.
There are a number of ways this is a smart play for Apple. The company has been growing its share of small businesses, and having a suite of tools to help those businesses manage devices and software is a big part of that. And it’s even smarter than it might look, because the goal is to make Apple devices the default option at work. Once your company email, device management, and employee directory all run through Apple, leaving gets a lot harder.
All of that is actually helpful context for the new thing Apple announced: ads are coming to Apple Maps this summer in the US and Canada. That means that local businesses with a physical location can claim their listing through Apple Business, upload photos, set a budget, and run a campaign. Apple’s automated system runs an auction and delivers ads, much in the same way ads work on Google or Instagram. Advertisers only pay when someone views or taps their ad.
What’s most notable, however, is how Apple is approaching ads in its Maps app differently from Google Maps. The company says it will only show one ad per search, and won’t allow competitor search targeting, meaning you can’t buy an ad that shows up when someone searches for the coffee shop down the street.
“A user’s location and the ads they see and interact with in Maps are not associated with a user’s Apple Account,” Apple said in its announcement. “Personal data stays on a user’s device, is not collected or stored by Apple, and is not shared with third parties.”
That’s a distinction that means a lot because, generally, consumers aren’t huge fans of ads. Partly that’s because they’re usually just in the way of finding whatever you were actually looking for, and partly because where there are ads, there’s usually data tracking.
Apple’s model is meaningfully different from how Google Maps advertising works. If you’ve opened Google Maps lately and searched for a local business, it’s not uncommon to see several sponsored results for other businesses before you see the one you actually searched for. I get that there’s money there, but it’s an absolutely garbage experience. If I type in the name of a business, there should never be paid results before the thing I asked you to show me.
Source: Inc.
NB: "It’s Apple’s Riskiest Bet Yet" No, it's not.…
Surveying multiple sources
We have a firm hold on the number of active devices as well as the number of iPhones in use. This is due to Apple reporting the data. The latest active device count was 2.5 billion and the estimate for active iPhones is 1.6 billion.
This patters also allows us to forecast the active base pretty accurately as the growth is predicted by the tight fitting linear approximation.
For more confidence it would be useful to confirm these predictions with different estimates. For this I turned to two additional sources: the Statcounter GlobalStats measure of Mobile Operating System Market Shares Worldwide coupled with Number of Smartphone Users Worldwide (forecast by Ericsson) and the Number of of iPhone Users from Bankmycell.
The total Smartphone Users history and forecast is shown below:
The Mobile Operating System Market Share Worldwide is shown below:
Multiplying these two yields the Mobile Operating System Users
The iOS users (orange line) can now be compared to Apple's own reported active device total (grey line) and the Bankmycell estimate as well as a derived estimate from Counterpoint (24% share of install base)
The graph includes forecasts from 2025 to 2029.
That, coupled with value captured per customer will determine the value of the firm.
It’s not unusual to see articles consisting of lists of reasons why Apple is facing an existential crisis. Agonizing and hand-wringing about Apple been a meme for decades, so much so that it had its own web site, The Apple Death Knell counter and the "Pray" for Apple imagery was a thing since the 90s. The original article from June 1997 offered 101 tips on saving the company starting with outsourcing hardware or scrapping it entirely.
Although the imminence of death has always been persistent, what’s changed over the years were the causes. There are far too many to recall or enumerate but I have through personal recollection and pattern recognition grouped them roughly into three categories:
For instance, during the last Apple Investor Event, I highlighted a typical article citing a list of problems for Apple. It was Titled “5 problems facing Apple it won’t talk about at is April event”. Published in August 2024 it cited the following:
When I read this I thought it must have been AI generated (hallucinated more likely). The dates did not make sense (August publication predicting an April event) and the competition cited was obsolete (Blackberry? Fujitsu?). It was a random list of Apple news framed as problems. Rather than try to refute them, I realized that it did not matter. The objective of the author was not to illuminate but to discharge something, anything, as long as it was negative.
Since last August a lot has happened. A new US administration, a new set of problems. Judging by the share price and news flow, Apple bashing is once again in vogue. (Incidentally, the stock price reached an all-time-high in December 2024 at nearly $260/share, a substantial increase from $170 or so six months earlier.) Today everyone and their editor is jumping into the scrum writing lists of Apple problems.
The pattern however seems a bit different this time. Almost all the catastrophes Apple has are externalities. Let’s try to enumerate them (inspired by Bloomberg.)
What these examples seem to share is that, except for the AI execution, they are externalities and thus of the third category. They are therefore problems related to being too successful.
Tariffs are a problem because Apple relies on massive scale and thus requires integrated production which creates concentration in production centers which are difficult to redeploy. Antitrust is a problem because Apple is too big and thus dominates developer distribution. Lawsuits center on having too much control—related to antitrust. Google antitrust remedy forces a decline in distribution revenue for browser search—a predicament coming about Apple’s enormous leverage vis-a-vis Google. Being so big means there is only one way to go: down, hence Chinese competitors gaining share. Lastly, AI execution problem is more Category 2 problem but would not be an issue if Apple did not insist on Privacy for its customers, something which caused much of its success with customers that it got too big.
Incidentally, profitability is now higher than ever, top line growth is consistent, bottom-line growth is even faster, the company is guiding growth in-line with current trajectory, the number of customers is at a new record, churn into the platforms continues, Apple is winning against non-consumption, capital is pouring into shareholder pockets and R&D spending is soaring. Cue the charts.
What the casual observer might note missing from the Doomed listicles is any mention of customer satisfaction, loyalty, acquisition and retention. The things the company actually talks about on earnings conference calls.
Do these externalities of tariffs, antitrust remedies, distribution dissatisfaction and non-iOS competition make any impact on how customers perceive the company?
We'll discuss this in the next few posts, as the lead up to presentations at ACTIVE: The Apple Investor Conference on Tuesday June 10th, the day after the WWDC25 keynote.
In yesterday's post, I proposed a way to inspect the Google-Apple distribution deal. It could be understood as a flat-rate deal where default placement of Google search in Safari is paid for with a constant run rate. This was the assumption for a few years with estimates for this run rate ranging widely.
Alternatively, it could be understood as a commission rate where payments are in proportion to ad sales. This is the 36% commission rate that has leaked during the antitrust trial.
Finally, it could also be seen as an access fee for each individual. This is the 17c/user/day that seems apparent from my calculations.
If one were to critique these options, I would disqualify the flat rate as unfavorable to Apple since the user base grows and behaviors can shift. The distribution commission sounds ideal but it would be problematic to enforce. How exactly can one track the precise income obtained by Google for each use of search, maps or YouTube on iOS or MacOS devices reliably enough with an audit trail for Apple to inspect?
For these reasons, I believe the fee of $x/user (or possibly active device as a proxy) makes more sense. The burden would be on Apple to prove that they have such active users/devices while Google could sample their metrics to confirm. Data on active users and devices are figures that Apple already supplies publicly and a method for determining them likely could stand scrutiny by outsiders. The server logs showing active devices can't be too onerous to share.
And so that's the model I'm going with. Having said that, the formulation of compensation for what is a complex distribution deal is likely complex. The abstraction to users/devices is one we have to live with given the challenges of inspecting the invisible.
That's the model of the present. The future however is clouded by the potential interference of authorities. In this case, an antitrust prosecution of Google is asking whether the way Google buys distribution is anti-competitive. The legal arguments are beyond scope here but the gist of it is that buying exclusive distribution when you are dominant may close options to not-so-dominant competitors.
The consequences, if we are to believe the commentary, are that the deal with Apple could be declared illegal and thus, stay with me here, Apple would lose a lucrative source of income.
Let me repeat this for emphasis: If Google were found guilty of anti-competitive behavior, Apple would be punished.
Let me expand for emphasis: If Google were found guilty of anti-competitive behavior, Google will be forced spend far less and be marched off a gangplank to swim in far greater profits.
Given this commentary one wonders why Google's management and legal team aren't begging the judge to rule against them and why Google shareholders aren't hoisting the prosecuting team on their shoulders.
The deal between Google and Apple exists only because both Google and Apple find it to be a better alternative to not having a deal or having any another deal. It does not exist because Apple benefits and Google doesn't or that Google benefits and Apple doesn't.
As I've explained the first post in this series, Google's benefit is distribution or access to 1.3 billion very valuable customers. For this they pay quite a lot and Apple receives quite a lot in turn. The relationship is shown in this illustration:
But—and here you should pay close attention—Apple also pays something for that benefit. Apple's $26b+ income from Google's TAC payments are in exchange for offering access. To Google. Not to someone else. And not to itself. Apple's costs are opportunity costs.
If Google had not offered the area marked with a question mark in the image above then Apple would have found alternatives. Alternatives that it's saying no to today.
Apple's alternative to the deal is what needs to be studied.
Could, for example, Apple have signed Microsoft for Bing access in Safari? Sure! They probably didn't because, as they declared, the quality of Bing would not be as good and as a result most people would switch out of it and thus result in lower revenues. Then again, perhaps Apple could make a fortune from Microsoft. Apple could have enabled Microsoft to rise to 40% market share in search, and also improving its quality. How much would that be worth to Microsoft? How much additional opportunity could Microsoft create from establishing such a strong consumer-facing franchise—something they have struggled to obtain even with vast expenditures. Surely Microsoft could be convinced to split revenues 50/50 for what would be all marginally zero-cost new upside!
Could Apple have developed its own search? Sure! They probably already did. Internet Search is a lot easier in 2023 than it was in 2007. Costs for computation, communication and storage drop exponentially while the number of internet hosts has stopped growing. Apple uses indexing in Siri and each iPhone, Mac and iPad runs a local search engine (Spotlight). Again, there are reasons why this is not as good or easy as signing Google up. But on the other hand, perhaps Apple could be making a killing on search. Why accept a 36% commission from Google when they could take 100% of 40% of all Search?
But most importantly, a declaration that this deal is invalid simply means that Apple and Google would craft another deal that would work around the restrictions. And that new deal would probably turn out to be more beneficial to Apple.
Why, I hear you ask?
Because Google would ask for the changes! Google would be forced to re-negotiate because they lost at trial. If that were the case, Apple would naturally ask for concessions. The most certain outcome would be that Apple will be paid even more for access to its users.
How much more? Remember that Google must ensure access to Apple's customers. According to my calculations Apple customers deliver almost 40% of its advertising revenues. The 36% commission suggests that there is plenty of head-room. Apple could ask for 50% and meter the results on a different basis. This would be a 39% increase in pricing for distribution.
This would result in a deal worth $41 billion/yr.
To implement workarounds, the companies would need to bridge their servers in order to ensure metering along new metrics. There would a great deal of complexity. But it would be worth it because both would find such a deal better than alternatives.
Fundamentally, the relationship between Google and Apple exists and will continue to exist because it is mutually beneficial and a better alternative to any other agreement or no agreement. If the current deal is found unsound, for whatever reason, a new deal will be created because it will need to be created.
Apple sits in a better position today than it ever did. Its customer base is growing while its pricing power with consumers is growing. In contrast, Google is the process of explaining that it did not break any laws. This while losing Android users to Apple. How does this lead to a vulnerability for Apple?
What we are witness to here is Apple increasing its distribution pricing power while simultaneously decreasing its opportunity costs.
The idea that Google being found guilty of anti-competitive behavior would result in a "hit to Apple earnings" is preposterous and fails on first inspection. The exact opposite is the likely outcome.
