Apple Intelligence

You could say that Apple Intelligence takes a handful of ChatGPT-3 chatbot class features, exposes each with a button to press rather than a text box to type in, and weaves in some pretty animations and colourful effects when you invoke them. That’s a curt summary, but it is not wrong necessarily. Everything Apple Intelligence does, we’ve seen before.

However, what makes it profound is the intentionality of the design, and the way in which these features are being realised. The marketing is straightforward and easy for people to understand, and the features are integrated naturally into the operating system surfaces that people already use. In fact, most of the ‘new’ features are things that the OS already ostensibly does; things like text editing and manipulation, notification management, smart replies, transcriptions, and — yes — emojis. Apple isn’t trying to convince people on wholesale new dimensions of what a phone is capable of. It’s taking what users already do, but made better by using modern AI techniques, so that users can extract more value out the other end.

When watching the demos and looking at the screenshots, all of these new buttons and panels fit in naturally as a component of the iPhone experience, as we know it. That’s a really significant point and speaks to how well they have executed here. They could have very easily screwed this up and announced a bunch of whizz-bang, discombobulated, AI-powered stuff that would ultimately land as extraneous gimmicks, go unused in practice, and feel like they were bolted on just for the sake of it.

The exception is the Image Playground stuff. I don't think it is completely pointless, but it's the hardest to find long-lasting value in. It might just be a rarely used gimmick, like Memoji.

I don’t think that’s the case at all. These features are compelling and — assuming they work as advertised — will be used en masse. I was pleasantly surprised that Apple is offering all of this for free. Despite being powered in part by the Private Compute Cloud server infrastructure, they have not used it as an opportunity to slap another subscription fee upsell onto its user base (at least not yet).

That means every iPhone 16 buyer this fall is going to be delighted by notification summaries of group text threads when they wake up in the morning, appreciate that when they scan their list of emails they can actually get a sense of what each message is about before tapping through, and enjoy creating emojis that have never existed before and sending them to their friends.

I am personally looking forward to all the new Siri improvements, although it remains a little murky as to exactly what will get better. The semantic index stuff isn’t shipping until next year, and it doesn’t seem to cover everything. New Siri might be able to dig up information from my emails and text messages, but will it still get confused if I ask it to turn off two HomeKit lights at the same time? I don’t know yet if Siri’s general knowledge base and understanding of user intent is evolved. I hope so, but Apple was evasive on specifics. My impression is that while new Siri makes progress on that front, paths to “I found this on the web” answers will still be a relatively common occurrence.

Perhaps my biggest disappointment of the entire endeavour is there is no indication as to how any of this could conceivably come to products like the Watch or HomePod, Apple’s most voice-oriented devices. Maybe they are working on solutions behind the scenes, but as it stands right now, it’s a big gap in the Apple Intelligence strategy.

What's New Is Thin Again

It has been a while since Apple introduced a new product that was markedly thinner than its predecessor, such that it was a selling point of the device. Throughout the early 2010s, it was almost a meme that every new shiny Apple gadget was touted as thinner. Apple rolled out the thinnest ever marketing blurb for the iPhone 4, the iPhone 5, the iPhone 6, the iPad 2, the iPad Air, the 12-inch MacBook and many more.

Those priorities shifted by the end of the decade, and certainly in the years since Jony Ive’s departure. On the phone side, goals to pack in bigger batteries and better cameras dominated any desires for more compactness. iPhone thinness peaked with the iPhone 6 at 6.9 mm in 2014, and has gradually increased ever since. The iPhone 15 Pro series measures 8.25 mm, in contrast. On the Mac front, the tribulations of the butterfly keyboard and related dramas resulted in a substantial change of course and the 2021 MacBook Pro was celebrated for its thicker enclosure, with customers grateful that Apple had not made compromises for thinness’s sake.

The 2024 iPad Pro bucks the trend, in style. Apple is incredibly proud to present the new iPad lineup as the thinnest Apple devices ever, and having seen the new models in person, it deserves all the plaudits. The new iPad Pro is breathtaking to hold in your hand. It is slender, and an insanely cool sight to behold. Picking it up feels like a flourish in itself. They also achieved all this whilst also improving burst and sustained performance of the device, thanks to the efficiency of the M4 chip and some clever heat dissipation trickery. The hardware is impressive and hard to fault.

Is this a one-off event or is thinness back in fashion again? It certainly feels like the new iPad is existence proof that Apple could make a really thin yet uncompromised MacBook, but we aren’t hearing that is on the cards quite yet. Interestingly, as of this week, there are rumours that Apple is readying an upscale iPhone redesign for 2025, that The Information describes as thinner than current models. I struggle to imagine that a thinner pro iPhone will have the same visceral impact as the iPad Pro though; the large protruding camera bump is a figurative and literal thorn in its side.

Apple Rumoured To Land Rights For 2025 FIFA Club World Cup Tournament

The New York Times:

FIFA, soccer’s global governing body, is close to an agreement with Apple that would give the tech company worldwide television rights for a major new tournament, a monthlong, World Cup-style competition for top teams that will be played for the first time in the United States next summer.

The value of the deal might be as little as a quarter of the $4 billion FIFA had first estimated, the people said. It is unclear if the deal with Apple will include any free-to-air rights, meaning the entire event could be available only to subscribers of Apple TV+, a factor over which senior executives at FIFA have raised concerns.

In his Puck newsletter, John Ourand said that Apple is garnering a reputation of being a ‘bottom feeder’ in sports. The perception is that Apple is an insignificant player, and can only secure the leftover scraps of sports rights, while the likes of Google, Amazon and — to a lesser extent — Netflix fight for the big national packages from the NBA, NFL and others. This is on a backdrop of the reality that Apple’s streaming service, Apple TV+, can only boast low two-digit millions of subscriber membership.

As you might expect, Apple would probably reject the idea of weakness and frame their position differently. Eddy Cue and team clearly value exclusive, unrestricted, streaming rights that are exploitable in many countries and regions. That’s what they got with the MLS deal and MLB Friday Night Baseball, and that’s what the FIFA Club World Cup would be.

Evidently, Apple has the stature and the financial might to secure deals of that nature, when they are available. The simple fact of life right now is the number of those possible deals are thin and far between. The major leagues are not yet willing to yield everything over to streaming, as they try and preserve as much of the reach and exposure that the slowly-dying-but-still-huge traditional broadcast networks provide.

With that context, it becomes more of a strategy question. Is it a blunder not to get what you can, even if it isn’t exactly what you want? Apple certainly has the confidence to sit out and taking things slowly. We’ve seen this play out, almost in the open given the amount of coverage it got, with the NFL Sunday Ticket negotiations. Apple was the frontrunner, but seemingly backed out when it didn’t go their way in the late stages of the dealmaking process.

Another similar situation transpired with the Pac-12. Apple was the leading bidder for a package of rights for the conference, with very similar terms to the MLS deal, but it was held up as some of the leading schools balked at the presented figures. Rather than overspend or give concessions, Apple held firm and they simply watched while member schools defected and the conference imploded altogether.

Perhaps it is misguided bravado, but I think Apple sees themselves as the ones in the position of strength, able to bide their time unless they can get what they want. That doesn’t sound like the mindset of a bottom feeder to me.

Apple Partnering With Google For Some iPhone AI Features


Apple Inc. is in talks to build Google’s Gemini artificial intelligence engine into the iPhone, according to people familiar with the situation, setting the stage for a blockbuster agreement that would shake up the AI industry.

Apple is preparing new capabilities as part of iOS 18 — the next version of the iPhone operating system — based on its own AI models. But those enhancements will be focused on features that operate on its devices, rather than ones delivered via the cloud. So Apple is seeking a partner to do the heavy lifting of generative AI, including functions for creating images and writing essays based on simple prompts.

Apple Silicon was such a triumph that I think people now lean too heavily into the idea that Apple should design everything in-house, or otherwise they are doomed to fail. That is a misguided view. As well-resourced and technically proficient as Apple is, they are never going to be the best at everything; outsourcing and licensing some components inside your products is the usual and pragmatic route. You have to pick your battles.

Just look at what happened with the 5G cellular modems. Apple hates Qualcomm’s guts, and their hefty patent royalties, but they use them because they are the best. If Apple had waited until they had a competitive in-house modem, they still wouldn’t be shipping a 5G iPhone today.

I think the situation with generative AI is analogous. Apple is working on their own stuff to defend against the future as AI becomes increasingly more embedded in their products, but as of 2024, their home-grown technology is not ready to match the quality of OpenAI or Google’s models. Thus, they are licensing Gemini.

Some of Apple’s AI-powered features will be serviced by on-device neural networks, some will hit Apple’s servers only, and some will be shuttled off to a Gemini backend. I expect this to happen transparently to the user, depending on the task at hand. The routing and mix of suppliers may change over time, even. At least for the foreseeable future, a dependency on Google as a dumb pipe supplier of AI tooling is not a weakness.

App Store Versus The EU Commission, Take Two


Web Distribution, available with a software update later this spring, will let authorized developers distribute their iOS apps to EU users directly from a website owned by the developer. Apple will provide authorized developers access to APIs that facilitate the distribution of their apps from the web, integrate with system functionality, back up and restore users’ apps, and more. For details, visit Getting ready for Web Distribution in the EU.

In my original commentary from January, I said these policies should be seen as an opening gambit rather than a final word on the matter.

Surely driven by backchannel negotiations between Apple and the EU commission, Apple has rolled out significant amendments to these rules in the last couple of months. Today, it announced perhaps its largest concession yet.

The rules as originally written were specifically orchestrated by Apple to offer “sideloading” without offering sideloading. The rules prevented the obvious, straightforward, approach of letting a developer host a binary on their website for users to download directly to their device. Instead, A developer wanting to offer their software outside of the App Store was forced to partner with an intermediary, an alternative app marketplace in which to list their app, and then would have to somehow explain to their customers how to install said marketplace in order to install said app. It induced business relationships and a whole bunch of complexity that didn’t really have a justification to exist other than frustrating the process.

The new “Web Distribution” method makes all of that obsolete, and feels like a direct response to the criticisms raised by the likes of Spotify. Eligible developers will now be able to offer up their apps ‘directly’. Some Apple oversight and system permissions scare sheets remain, but the essence is Spotify will now be able to have a link on their website that initiates an install flow for the Spotify app. The obstacle of the viability of the Core Technology Fee remains, but that’s about it.

Whether Epic Games will be able to distribute Fortnite in this way is currently unclear. Apple’s rules say that a participating developer must be a “member of good standing in the Apple Developer Program” and “have an app that had more than one million first annual installs on iOS in the EU in the prior calendar year”. Unless Apple makes an exception, as is, Epic doesn’t meet either of those requirements.

The Launch Of Apple Vision Pro

I was bracing for the critical response and early reviews to the Vision Pro to be dire; a joke, a waste of a time. But the sentiment is actually far more positive than I expected. Amidst the ungainly ergonomics of the first-generation hardware, and the window management omissions of a first-generation operating system, people have found glimpses of greatness. More so for entertainment, but productivity use cases too.

Those touting daily usage will inevitably taper off, and I’m sure many will stop remembering to charge it altogether in about two months’ time. There’s just not enough you can do with it right now, with a dearth of worthwhile content to consume and slim pickings of high-quality apps to use. A portion of enthusiastic buyers will end up returning it as they confront the physical realities of something so heavy and bulbous sitting on your face. But if you squint enough, you can almost see the future. The day one Vision Pro experience already feels more capable than an iPad in many respects, and it’s a great plane computer. Those niches and bubbles of utility will only grow larger over time.

Apple doesn’t enter a new product category too often. With the iPhone, iPad and Watch, Apple was coming out swinging from the start with a product concept that was ready to draw huge mass market attention. The Vision Pro is obviously not going to do that, nor does Apple expect it to. The market dynamics of AR/VR have basically forced Apple’s hand into launching a bit earlier in the hardware development curve than usual. More than anything the company has released before, the Vision Pro bumps against the limits of the state of the art in so many ways.

The first iPhone does not look that different from the iPhone we know today; the bezels disappeared, the cameras got bigger, but it’s still unmistakably of the same family as the device Jobs pulled out of his pocket in January 2007. Same with the iPad, and the Watch. However, I expect that the Vision Pro’s tenth incarnation will be vastly different from what we have today. There’s just so much room for everything to get better, and for key tentpoles of the design to shift based on what becomes technically possible in the future.

App Store Versus The EU Commission


Apple is sharing new business terms available for developers’ apps in the EU. Developers can choose to adopt these new business terms or stay on Apple’s existing terms. Developers must adopt the new business terms for EU apps to use the new capabilities for alternative distribution or payment processing.

The new business terms for apps in the EU are necessary to support the DMA’s requirements for alternative distribution and payment processing. That includes a fee structure that reflects the many ways Apple creates value for developers’ businesses — including App Store distribution and discovery, the App Store’s secure payment processing and related commerce services, Apple’s trusted and secure mobile platform, and all the tools and technology to build and share innovative apps with users around the world.

Firstly, I do not consider the policies introduced this week as the end of the conversation. In many ways, it is the opening gambit. The Digital Markets Act leaves much to interpretation and the EU regulators have essentially left it up to companies to apply the law, and then assess whether the changes that companies implement is sufficient. I view the published Apple rules as a working proposal, very much subject to change. Apple would rather it wasn’t weakened from here, of course, but I think the EU has shown that it is operating proactively and will follow up with addendums of legislation and enforcement action, as it sees fit.

As it stands today, here’s what iOS developers inside the European Union are facing. They can either keep the status quo, stay on the App Store and pay the traditional 85%/15% and 70%/30% commission split on digital purchases, or choose to adopt so-called “alternative business terms”.

These terms deconstruct the commission structure into two components and instate a new Core Technology Fee; an annual levy payable on a per-install basis. The commission is actually less. Apps listed in the App Store are subject to 10% commission for small developers, or 17% for larger developers (those that exceed $1 million in annual revenue). That’s down from 15% and 30% respectively.

However, for use of Apple’s In-App Purchase payment system, there’s an additional 3% fee. So, really, it is 13% or 20% to deliver the same end-to-end store experience.

Developers have the flexibility to use alternative payment systems, in which case the 3% is not due. If they don’t want to use the Apple App Store, they can list their app in a third-party app store (Apple refers to these as an “app marketplace”) and not owe any commission at all.

So far, so good. Spotify, Epic, and the rest will never be fully satisfied, but if that was the terms, I think they’d come away very happy.

The snag is the Core Technology Fee, priced at €0.50. The CTF is paid the first time an app is downloaded (or updated) by a user in a twelve month period. The next twelve month period, the fee is due again. Apple has essentially instated an annual fee for every active user of an app, whether the developer is able to monetise that user or not. You could also argue it has wider scope, as active installs does not correlate to active users necessarily. An iPhone user may stop using an app altogether and simply forget to delete it from their device. But it will keep receiving automatic updates in the background, and that will invoke the CTF, even if they never opened the app themselves. A downloaded app that is never launched a single time results in an annual bill to its maker.

The impact of the CTF on the cost-benefit calculus depends on the popularity of the application. For smaller developers, it’s actually not too bad, at least with a surface level examination. Apple gives you a million free installs before the CTF comes into effect. That’s a sizeable figure and many apps will easily stay inside that threshold, even accounting for some bloating of the numbers due to inactive users and whatnot.

But, Apple stacked the deck well. The upside is relatively small. Assuming you are switching away from the 85%/15% split of the Small Business Program, you’ll now be paying 13% commission (10% App Store + 3% IAP). That’s only a saving of just 2% compared to the status quo. In exchange, you bear a huge downside risk of CTF fees if your app grows and crosses the million install mark. You are basically betting your app will never get big, for the entire lifetime of your company, in exchange for 2% extra revenue. That doesn’t sound like a very compelling proposition. You might argue that taking this route allows you to set up a shop inside an alternative app store, and forgo Apple’s commission altogether. That is true, but smaller developers are unlikely to have the marketing budget and inertia pull to beat the might of distribution through the Apple App Store.

For bigger developers, the costs of the CTF are significant, but perhaps surmountable in some cases. I mean, the costs are immense. Averaging 50 cents in revenue per user, including a ‘user’ that may never open your app at all, is really hard. If you have 2 million installs per year, that equates out to almost a half a million euros in CTF fees alone, notwithstanding App Store commission or any other marketplace charges.

Apps supported by advertising will never take these terms. Why? They can live for free in the App Store. In the 70%/30% model, Apple takes no money earned from advertising so these apps pay nothing to Apple at all. If they want to venture out, they are slapped with a 50 cent per user fee. They’ll never take it.

Freemium apps, like games with micro-transactions, are also going to struggle. A free-to-download game with ten million users has got to find €4.5 million to give to Apple each year, solely for existing. You are going to need some big money spender whales to offset the costs of all the people that pay nothing, to have a chance at breaking even.

Apple remarks that it estimates less than 1% of developers would pay any CTF fees, because most apps don’t clear a million annual installs in the EU. But it’s the 1% that are really driving these changes. It’s precisely that 1% that lobbies governments and motivates the creation of the Digital Markets Act in the first place. Spotify, Epic, Netflix, Tinder, etcetera.

These ilk of companies do have high average-revenue-per-user, and I think probably could afford the CTF rates. But they aren’t going to be happy about it. They will argue why do they have to pay Apple anything. It’s almost ironic that Spotify has been pushing for Europe to crack down on the App Store monopoly, and the offer on the table actually has them paying more to Apple than they do now. Right now, Spotify pays zero. But Apple’s new terms sees Spotify paying 50 cents per user per year to them, if they leave the App Store and make themselves available elsewhere.

The other big gotcha here is that iPhone app sideloading is not a free for all. Spotify cannot simple host an app binary on its website, like Mac apps can. You can only install Spotify through an app marketplace. That means Spotify has to educate users to install the alternative app store that Spotify is contained in, then find and install the Spotify app from that store. It also implicitly entails Spotify into some kind of business arrangement with this third-party app store, possibly involving some form of commission revenue sharing on top of the mandatory Apple CTF. By the way, Apple also imposes the CTF on the app marketplaces themselves, which conveniently prevents these alt-stores from being open season charities. By necessity, the marketplace will need their own intermediary business model to at least afford the Apple fees for every user that installs them.

My current prediction is that almost nobody will actually make the leap of faith into the “alternative business terms” as they are written. Small developers don’t have the leverage to experiment and will stay put. Out of the big companies, Fortnite might be the only one, with Epic seemingly insistent on launching an iOS Epic Games Store, although that may not be viable in a 50 cent CTF world. Spotify, Netflix and the other constituent members of the aforementioned 1% will remain in the App Store, and appeal to the EU to make Apple make more changes. These companies will push for the abolishment of the CTF altogether; only time will tell if the EU commission is receptive to that.

Apple Says It Will Not Make A New 27-inch iMac

The Verge:

Apple will not be making an Apple Silicon version of the 27-inch iMac to replace the Intel-equipped model that it discontinued in 2022. The company is instead focusing the iMac line around the 24-inch model that was first released in early 2021 and just updated with the new M3 processor this fall.

Apple PR representative Starlayne Meza confirmed the company’s plans to The Verge. The company encourages those who have been holding out hope for a larger iMac to consider the Studio Display and Mac Studio or Mac Mini, which pair a 27-inch 5K screen with a separate computer, compared to the all-in-one design of the iMac.

The all-in-one PC is perhaps the epitome of Apple design principles. The form factor pushes elegance and simplicity to the extreme. It accepts that some tradeoffs are necessary, and rejects trying to satisfy every little edge case demand, in order achieve the ultimate outcome for a certain mainstream customer. The end result is an iMac, a really great desktop computer. I am not the biggest fan of the aesthetics of the 2021 iteration, but that is beside the point. The iMac is great for what it is.

However, the market dynamics are not in the iMac’s favour. Portable laptops dominate everything these days, and frankly Apple Silicon has minimised many of the traditional disadvantages of choosing a laptop versus a desktop.

The iMac is not a dead end product line, but it naturally warrants far less company resources than the MacBook Air or MacBook Pro. I think we can all observe that the iMac is on a development cadence that sees it receive significant changes once or maybe twice a decade, subsequently upgraded with new more powerful chips every one to two years. The 2021 iMac was the big uplift, the 2023 revision is the example of the spec bump.

The base iMac is clearly set as is, with a 24-inch inbetweener design meant to split the difference between the previous Intel lineup of 21.5-inch and 27-inch sizes. That’s what Apple’s on-the-record press statement is affirming. But just like Apple offers both the Mac mini and Mac Studio, I do think it would be a shame if the company can never justify making a higher end iMac ever again. A true Apple Silicon successor to the iMac Pro could be very compelling, even if not a big seller.

It would make sense for this hypothetical machine to sport an even larger screen size, perhaps 32 inches to rival the Pro Display XDR. To further differentiate from a Studio Display setup, the screen could feature significantly higher display resolutions at 120Hz ProMotion frame rates, taking advantage of the fact that the bandwidth of an internally integrated display controller can far exceed what is supported by external HDMI or DisplayPort cables. The chassis would obviously be designed to house the hungrier Max/Ultra Apple Silicon chips, and — going out on a limb — I’d hope they’d choose an industrial design that can remove the chin bezel underneath the display for good.

Apple Announces October 30 Event, Held At 5 PM Pacific

Ever since the first virtual WWDC, I have not been enamoured by the virtual event format — it’s just less fun. In 2020, doing a pre-recorded video was a necessity. Now, it’s a choice. Google, Microsoft and others have done a few live presentations in the last year and it’s engaging, even if I don’t care so much about what they are announcing. In contrast, Apple’s events are lacking that feeling of flair and vitality. The novelty of the swooping Apple Park shots has worn off, and what’s left is something quite dry and quotidian. I’m sure Apple marketing loves controlling the message with a carefully-crafted and perfectly edited 90-minute video. I call it picking the dull, risk-free route.

Albeit still a pre-recorded video, I am hopeful this October 30 event will take the opportunity to mix it up a bit. ‘Scary fast’ and the dark imagery certainly suggests they are embracing the end-of-October Halloween spirit, which would be a fun twist on proceedings. It’s also the first ever Apple event held in the evening, kicking off at 5 PM Pacific (rather than the usual 10 AM start). Different is fun. It does mean that I will be watching it at midnight local time in the UK, but swings and roundabouts … I’ll take it. I’m assuming it will be under an hour in duration.

My expectations are proportionate. Bloomberg’s Mark Gurman is very confident that the M3 Apple Silicon chip generation is beginning, and I have no reason to doubt that. I expect to see the launch of the M3 24-inch iMac, and M3 Pro/Max MacBook Pro. In summary, notable spec bumps, same designs.

The A17 Pro Chip Naming

Perhaps one of the most provocative changes Apple announced at its event this month is the abandonment of the ‘Bionic’ moniker for its A-series chips in favour of a ‘Pro’ adjective. The conspicuously veiled truth in choosing that name is the absence, at least to date, of a non-pro A17 chip.

For the longest time, Apple would update all of its flagship iPhones with the same, new, chip each fall. Starting last year, Apple set out to distinguish the base mode and higher-end phones by only upgrading the chip in the iPhone 14 Pro and iPhone 14 Pro Max. They got the A16 Bionic, whereas the base model 14 and 14 Plus were powered by the A15 Bionic, the same chip as the year-ago iPhone 13 Pro. Apple repeated that product stratification strategy again, such that the base model 15’s house an A16 and the 15 Pro and Pro Max get the upgraded chip, the A17 Pro.

The ‘Pro’ chip naming is clearly an indicator that this pattern is here to say. One outcome is that next year, the iPhone 16 gets a plain ‘A17’ chip, perhaps merely a binned variant of the A17 Pro silicon, with one less CPU core or something insignificant. The Pro phones would continue ascending to be even more powerful, with A18 Pro innards. The non-pro A17 chip could also find its way into other lower-end iOS devices like the entry-level iPad, at some point.

The other way to look at this is to consider this a transitional year. Perhaps, the A17 will never make its way into a future iPhone and the A17 Pro is a one-off outlier. It would follow that next fall, Apple’s strategy would fully reveal itself by introducing both an A18, and an A18 Pro chip lineup, for the iPhone 16 and iPhone 16 Pro respectively. This would certainly provide naming symmetry across the models of the same year, and closely mirrors what Apple does on the Mac side, in which the M-series of chips comprise a family that includes higher-end (Pro, Max, Ultra) variants. You could even easily foresee an A18 chip above Pro, if the iPhone Ultra rumours come to pass.

How Many Subscribers Does Apple Have Exactly?


“We are happy to report that we had an all-time revenue record in Services during the June quarter, driven by over 1 billion paid subscriptions, and we saw continued strength in emerging markets thanks to robust sales of iPhone,” said Tim Cook, Apple’s CEO.

Apple’s hardware sales growth has been negligible-to-flat for several quarters in a row. More than ever, their quarterly financial statements depend on the Services business to show a record result. This quarter, Apple reported 8% revenue growth for Services, topping $21.2 billion for the quarter, and announced it had reached 1 billion paid subscriptions on its books. They also said the installed base of active devices hit an all-time record, without disclosing a specific figure. (Apple has repeatedly argued that a bigger install base means more customers will engage with services over time, and so far that has held up.)

But that’s about all Apple will tell us as to the performance of Services. It hasn’t reported Apple Music subscriber numbers since 2019, nor has it ever given hard figures about the performance of Apple TV+, Apple Arcade, News, iCloud, or Apple One in general. A billion subscribers is a huge headline figure, but it obscures the real story of what most people think of when you say ‘Apple services’. Services includes the App Store, and so a majority of that 1 billion total includes In-App Purchase subscriptions from third-party apps in the App Store. Although we never know for sure because Apple won’t tell us, it follows that the majority of Services revenue growth hails from the 15-30% commission Apple collects on those in-app purchase transactions.

If I was a financial investor, I would be growing increasingly dissatisfied with the murkiness of the Services business. For Apple’s flagship growth unit, it’s really hard to get a read on its performance. The golden goose of Apple’s stronghold on the App Store is constantly under threat from regulation, but we can’t measure the potential impact on Services revenue. The success of Apple’s content services are a hedge against the risk of App Store commission drying up, but we don’t know anything about the state of those offerings — we can’t even say for sure they are successful.

Apple stopped reporting unit sales numbers for its hardware products in 2018, but it still reports revenue per division. Rather than a single ‘Hardware’ revenue total, Apple reports quarterly revenue breakdowns for iPhone, Mac, iPad, and Wearables, which gives some visibility into how each product line is doing over time. In contrast, Services is completely opaque. There is no breakdown provided, just one total revenue figure. I am surprised there isn’t more pressure here from Wall Street for Apple to reveal more details. A few years ago, Services was small enough that it didn’t really make sense to split it out. These days, though, Services is so large that it is bigger than the Mac, iPad and Wearables units combined in revenue terms. If you hypothetically split out the App Store as a sub-unit, there’s a decent chance it alone would be larger than the iPad on the balance sheet.

At the very least, I think it’s time for Apple to be more transparent about that subscription total. How many of the 1 billion subscriptions are for Apple’s services, versus third-party subscriptions? And how many unique users does that 1 billion subscriptions represent? It’s not even clear to me how it is calculated. A user who pays for iCloud and Apple Music presumably counts as two subscriptions, but as a subscriber to Apple One Premier, do I count as one subscription, or six?

iOS 17 Dramatically Improves Autocorrect With Smarter Algorithms And Smarter UI

During the WWDC announcement, Apple focused on how the keyboard autocorrect system in iOS 17 is powered by machine learning based on ‘transformer’ neural networks, with the aim to enhance accuracy and make the corrections feel more personalised to each user.

Having used iOS 17 for a month so far, you can definitely feel the difference. The corrections are better. It feels like it knows what you meant to type far more than any previous version of the software. It also seems more resilient to typing slang. I noticed it can cope with common texting lingo reductions like ‘wut’, opting to leave them alone instead of insisting a correction to the nearest word it finds in the dictionary. In a very unscientific test, I tried typing ‘wut’ on an iOS 16 phone — and it kept changing it to ‘wit’. Overall, the iOS 17 engine is more useful and less obstructive.

But algorithm improvements are only part of the story. Obviously, it still won’t get it right all the time. But in those cases, the experience of managing autocorrect is also improved through a superior UI. When the system does make a mistake, it is far less punishing as the interface now gives you way to quickly revert autocorrect changes. As you type, any corrected words are briefly underlined in blue. This means you can more easily notice when autocorrect changed something, and address it immediately, instead of getting through your whole message and only then spotting an error. Tapping on the underlined word shows a popup menu that lets you undo to what you literally typed, as well as some alternative suggestions to pick from. Word predictions are also much more useful, showing inline as you type. Just hit the spacebar to accept the suggestion and keep typing your message.

The smarter algorithms and smarter UI come together in a very tangible way to offer a meaningfully better experience.

The Apple Vision Non-Pro Will Still Be Very Expensive

The Apple Vision Pro headset is wowing everyone who gets a chance to try it, with universally positive impressions being published from press who got to have the 30-minute demo experience. In those articles, I’ve noticed a common refrain that I don’t think is accurate. The presented idea is that whilst this particular model is arriving “early next year” as something too expensive for the ordinary person, Apple is already working on the second-generation non-Pro model, expected in 2025, and that will solve the price problem and make the headset accessible for all.

In short, I think that’s way too optimistic. The visionOS platform has long-term potential to be a mass consumer device, but we are talking long-term. The second-generation Apple Vision will be just as out of reach to most as the Vision Pro is today. With the Vision Pro at $3500, a stripped down cheaper model is still going to be very pricey. I’m anticipating somewhere around the $2000 mark. For comparison, that is around $1000 more than the most expensive Quest headset; Meta launched the Quest Pro as a $1500 device in October 2022 and swiftly dropped it down to $999 in March.

The Meta Quest Pro does not sell in volumes anywhere near levels that could be deemed ‘mass consumer’. Meta has sold tens of millions of units over its lifetime, but that stat is dominated by sales of the entry-level Quests, which retail under $500 — and mostly targets the semi-casual VR gaming market.

I’m not sure the Apple Vision product line will ever reach prices that low, at least as Apple how envisions it (pun intended) today as an augmented reality spatial computer. The EyeSight feature alone must add hundreds of dollars in cost to the bill of materials — between the curved lenticular front-facing OLED display and the sensors needed to drive it. Without considering anything else, the existence of EyeSight means the lowest I can ever see an Apple headset going is $1500 — and that’s not a near term thing, that’s many years off.

Putting aside the myriad other drawbacks of the device’s form factor given current technological constraints, the price alone means I am very bearish on the mass consumer prospects on the Apple Vision product line. I’d wager it will take at least three generations of hardware evolution to get to something appealing to the mainstream.

It took Apple one year to sell ten million iPhones; I wouldn’t expect Apple to achieve 10 million visionOS unit sales until 2027-2028. To reiterate, that’s a five year timescale. They are playing a very long game.

Final Cut Pro And Logic Pro Announced For iPad


Apple today unveiled Final Cut Pro and Logic Pro for iPad. Video and music creators can now unleash their creativity in new ways that are only possible on iPad. Final Cut Pro and Logic Pro for iPad bring all-new touch interfaces that allow users to enhance their workflows with the immediacy and intuitiveness of Multi-Touch. Final Cut Pro for iPad introduces a powerful set of tools for video creators to record, edit, finish, and share, all from one portable device. Logic Pro for iPad puts the power of professional music creation in the hands of the creator — no matter where they are — with a complete collection of sophisticated tools for songwriting, beat making, recording, editing, and mixing. Final Cut Pro and Logic Pro for iPad will be available on the App Store as subscriptions starting Tuesday, May 23.

It’s been a while since Apple has released software with such craft and care, as is on display here. Without even using the apps, the screenshots stand on their own as an impressive feat. I love how these apps are sophisticated in scope whilst still highly accommodating to touch input. A fair few ‘pro’ apps that have come to iPad in recent years just assume the user is working with an attached keyboard and mouse. They basically give up on the touchscreen part of the tablet form factor, because it’s easier to get their desktop app ported that way. No such shortcuts have been taken here. You could ably use Final Cut and Logic with just your finger on an iPad screen. I love to see it.

When these apps ship in a couple of weeks time, there will immediately be a laundry list of complaints from pro users about missing features and reasons why these iPad apps can’t replace their Mac workflow; many of those reasons will be the fault of the platform itself, like file management or access to plugins and I/O. Those negative headlines will inevitably happen, but I don’t think it matters much. This is Apple seriously putting its stake in the ground, and some people will be able to use these apps for real right out of the gate. Apple can and will keep chipping away at solving the outstanding problems to capture more and more use cases.

Apple has been crying wolf about the iPad as a productivity machine for far too long. You can’t deny that this announcement is a great start to fight that narrative.

Humane Unveils Its Wearable Device For The First Time

Humane, the secretive startup founded by ex-Apple software design chief Imran Chaudri, finally went public with Chaudri showing off their device for the first time at the TED conference last week. I’ve seen a recording of the 15-minute presentation, which unfortunately is yet to be officially published online.

Chaudri’s talk is centred on the premise that technology (mainly through the smartphone) has invaded all of our lives too much. The idea is that personalised artificial intelligence can be used to dramatically change how we interact with technology. Rather than proactively opening an app to do something, AI can be an ambient thing that is there when you need it, works in the background of your life, and mostly stays out of your way. To make this a reality, Humane is introducing a new product: a wearable that resembles a rectangular pin badge. Chaudri is wearing one on his jacket pocket during the presentation. He sets out the vision of their product as something that is “screenless, seamless and sensing”.

Chaudri demonstrates the unobtrusive utility of their device by asking it ‘Where can I find a gift for my wife before I have to leave tomorrow?’. The badge audibly responds with a suggestion of a nearby shopping district. It’s a cool demo in that it gives a more useful contextual reply than what we’d expect from a typical response from Siri or a similar voice assistant would give today to that same query, with the Humane system clearly being infused with large language model smarts. (Assuming the demo is legit and not scripted canned responses, of course).

However, I do not see how that demo justifies the form factor of a clip-on screenless badge. In fact, if I wanted to actually go to said shopping mall, I am left wanting a screen to visualise the area on the map and show me directions. Innately, then, I think that a phone with a smarter inbuilt assistant supersedes the Humane product, as does a smartwatch for that matter. Watches have screens but they are just as seamless and subtle as a wearable on the front of your jacket, I’d argue.

Humane’s counter to the visual information problem seems to be the inclusion of a short-throw projector in the badge. This allows the device to beam text and images to a nearby surface. If you are standing upright, away from a table, this means holding up your hand awkwardly in front of you so the badge can project stuff onto it. So, despite the screenless pitch, what they have essentially created is another screen after all; one with unstable reproduction (thanks to your naturally shaky arm/body), relatively low color fidelity and resolution, and uncomfortable ergonomics.

We don’t have much else to go on in terms of technical specification, but Chaudri did stress that the device is meant to be used wholly independently; ‘you don’t need to carry a phone anymore’ was not said, but certainly an implied notion. Maybe there’s room for another wearable accessory in our lives, and if Humane had positioned their product in that vein, I would be far less sceptical. I am not onboard with the presented vision that they are pioneering the primary future of personal computing.

Some Apple Employees Unconvinced By Headset's Purpose

The New York Times:

As the company prepares to introduce the headset in June, enthusiasm at Apple has given way to skepticism, said eight current and former employees, who requested anonymity because of Apple’s policies against speaking about future products. There are concerns about the device’s roughly $3,000 price, doubts about its utility and worries about its unproven market. That dissension has been a surprising change inside a company where employees have built devices — from the iPod to the Apple Watch — with the single-mindedness of a moon mission.

Apple’s recent history of new product category launches denote major, culturally impactful, events: AirPods, Apple Watch, iPad, and of course — the pinnacle of the bunch — the iPhone. The Watch didn’t set the world on fire immediately, but I think it still qualifies to be among the bunch, racking up tens of millions of sales within two years. AirPods took a while to ramp up too.

The rumoured headset product is simply not going to meet that same level of appeal, but is that a necessarily a bad thing? It depends. If Apple presents it as the next big thing, laden with superlatives, then it will backfire. If it approaches its introduction in a more subdued fashion, in a similar vein to the launch of something like the Pro Display XDR, then it’s probably fine. Don’t set unrealistic expectations and people will not be disappointed. Apple Reality Pro is the start of a long journey, and the billions of dollars of research and development will eventually culminate in something monumental. Just not yet.

Being active in the augmented reality space is clearly important, and Apple has signalled as much with six major versions of ARKit under its belt already. The v1.0 headset hardware is the next step on the journey. In the best case, it will establish Apple as a market leader in the space, evangelise to developers and kickstart an ecosystem. It might make some inroads in the enterprise. I don’t think many ordinary people will get on board. The rumoured second-generation model has a better shot, but even that will probably be priced above $1000. But maybe by the time that second-gen comes out, Apple and developers noodling with the first-gen might have figured out some killer app use cases that will allow that price to be somewhat justified. If not, no biggie. There’s always the third-gen. And after that ships, the state of the art of technology will hopefully be closing in on making the ideal form factor — thin and light glasses — viable. And when that happens, Apple will be ready.

Eddy Cue once said, if Apple only did things that were as big as the iPhone, they would never release another product. Back when Apple was far less flush with cash, it had to make an immediate splash to survive, let alone thrive. Nowadays, the company doesn’t have that pressure, and it’s implausible that it could one-up itself forever. Sometimes it makes sense to start small.