Now we know that June 9 is the day Apple will kick off WWDC 2025. I’ve been covering WWDC since before Steve Jobs came back to Apple, and this year is shaping up to potentially be the most interesting and certainly most dramatic version of Apple’s most important event of the year.
If you didn’t know already, WWDC (which stands for Worldwide Developers Conference) is important for more than just developers. It’s literally Apple’s New Year’s Day, the day that the company rolls out its plans for all of its platforms for the next year. With the impending release of iOS 18.4, macOS 15.4, and the rest, we are at the tail end of last year’s cycle. Work at Apple is now shifting even more to the stuff we’ll be using over the next year, beginning with the announcement of new features and new decisions on June 9.
Last year’s event, which featured the rollout of Apple Intelligence–including, yes, some features that never ended up shipping–was certainly dramatic. But at the same time, we all pretty much knew what was coming: Apple was desperate to be seen as a player in the AI game, and so it was going to blow that horn as loud as it could. This year’s fascination is more subtle: What now? Doubling down on Apple Intelligence? Apologizing for last year’s, er, overexuberant promises? Changing direction? Staying the course? The last year has flipped the table on Apple’s usually conservative and careful platform-building plans. This year’s possibilities are wide open.
The elephant in the room
Last year’s rollout of Apple Intelligence wasn’t subtle, and by most accounts, it was a rush job after the company realized it was behind in the LLM game. This last year has been bumpy, with some features working okay, others embarrassing Apple, and of course, some not showing up at all. It’s never been more evident that Apple is not a company that is capable of rushing operating-system features into production in a matter of a few months. (Maybe it should be, but that’s a different topic.)
Apple felt the pressure to jump on the AI bandwagon and it’s been a bumpy ride ever since.
Apple
But it’s been more than a year since that fateful decision was made, nearly ten months since WWDC 2024, and five months since Apple Intelligence began shipping to customers. So, what has Apple learned from this entire experience? This is the key question, and while the company won’t tell us the answer directly, we should all watch closely to see how it alters its approach.
My guess is that Apple will consider its primary goal of WWDC 2024–to create broad knowledge of “Apple Intelligence” as a brand name that makes Apple vaguely seem in the AI game–to have been achieved. Those who are knowledgeable about the current state of AI will roll their eyes and cite dozens of ways in which Apple is nowhere near the cutting edge of AI, and they’re not wrong. But I think that in the broadest perception of Apple as a company, consumers are vaguely aware that Apple says it has AI. And even people closer to the tech industry have to admit that Apple seems to have a huge desire to get into the AI game, even if it’s stumbled out of the blocks.
That’s why this coming year, Apple Intelligence will probably shift to more of a clean-up and consolidation phase. The last year was messy. Bits of Apple Intelligence were jammed into places where they didn’t quite fit or make sense. I hope that Apple will refine the existing tools, ship updated (and hopefully much improved!) versions of its own models, and also focus on shipping the stuff it promised last year that it couldn’t deliver this year–if it can. If it can’t, if those features prove to be elusive, that will continue to be a black eye for Apple–but it needs to own it and move on.
As a result, I expect that Apple will only announce a small number of new Apple Intelligence-derived features, not the fusillade of features we got last time. After a year, I hope that the company will understand better what it’s capable of accomplishing in a single year.
WWDC 2025 needs to prioritize the needs of developers.
Apple
What about developers?
Lost in all the Apple Intelligence hype last year was the somewhat astonishing fact that at its developer conference, Apple unveiled a load of AI features that offered almost no way for app developers to get in the game. The biggest story for developers at WWDC 2024 was App Intents, which would allow apps to better integrate with all the features that Apple has now admitted it can’t ship. So in hindsight, WWDC 2024 had almost nothing for developers at all.
This needs to change, and WWDC 2025 is the place for it to happen. App developers should view Apple platforms as the best places to build AI-enhanced apps. That’s good for Apple. The company has invested a lot in creating hardware that’s pretty solid for running AI applications (thanks, Neural Engine). The next step should be offering developers access to Apple-blessed ML models on its devices, and easy systemwide integration with external models from Apple and third parties. Apple built Apple Intelligence for itself; it’s time to turn the tables and let the app developers start innovating.
Speaking of external models, Apple has spent a year with a single integration–to ChatGPT. It’s a great model that’s very useful for lots of stuff, but it’s not without serious competition. Apple should announce that it’s integrating many more models, letting users choose their favorite or allowing the operating system to pick the right tool for various jobs. The company also needs to get over its overwhelming fear of users using these tools–the current implementation of ChatGPT integration requires multiple warning steps unless you dig into settings and shut them all off.
By all means, Apple needs to give users the ability to opt out of all AI queries and content. But given how broken Siri is these days, it might also want to make it much easier to opt in and kick most queries to remote sources. Until Apple has a chance to beef up Siri, the Siri agent should only offer to perform the most basic functions and kick everything else out to an LLM. It’s a quick fix, even if it will require Apple to swallow a little pride.
Apple needs to allow the integration of other AI models besides ChatGPT
.
Foundry
And then there’s Siri
Apple just rearranged the executive ranks and put Mike Rockwell in charge of Siri. If you’re hoping for major Siri announcements at WWDC 2025, though, I have a stern warning.
We all want Siri to be fixed. But if Rockwell is just now being put in charge, it means that any announcements for Siri fixes in the next year would probably be just as rushed as the Apple Intelligence features were last year! As a result, I expect that Apple will be a little less aggressive on Siri, while still making some promises for early in 2026. As I detailed a few paragraphs ago, perhaps the best solution is to turn more Siri queries over to other LLMs in the meantime just to stop the bleeding.
I realize the temptation to make big claims about changes to Siri will be great. But… we were just here last year. I hope Rockwell rights the ship, and Siri gets in gear, but I want to see Apple show that it’s learned its lesson about overpromising and underdelivering.
A new design? Sure!
There have been lots of recent reports that this year’s OS updates will feature a new design language. It sounds good to me. Here’s how the cycle goes: A company releases a major new redesign. It pushes things a bit too far in the interest of making change. Over the course of a few years, the most extreme stuff gets sanded off, and everyone settles in. Then many years pass, and new doodads get added here and there as one-off additions until a decade passes and your clean new design is now a monstrosity. It’s time for a new, consistent design.
While a lot of people are reporting that the new design language is “based on visionOS,” I don’t think that’s the case. My guess is that Apple has been spending years working on this design, and it’s been built from the ground up to work across all of its devices… including, logically, the new Vision Pro. But Vision Pro shipped before the new design was ready to roll out elsewhere, so it became a testbed for some of the ideas of the new design. As a result, I expect aspects of this design to be reminiscent of visionOS–but they’re all from a common source, not just being copied from visionOS.
I do think that it’s time for a fresh design, and for Apple to extend its design philosophy across all of its operating systems. While a lot of Mac users might cringe at the idea that the Mac is going to just end up looking like an iPad, I actually think that this process should be a good thing for Mac users. A good, thoughtful design system should take into account the things that make all of Apple’s devices unique–while bearing a family resemblance and working similarly, like they’re all from the same company. I would argue that a lot of the most glaring design sins in macOS in the past few years have come from iOS designs just getting sprayed into macOS without a whole lot of thought. If Apple’s designers have done their jobs, this time around, the Mac will be considered along with iOS, and they’ll make the right decisions for each of Apple’s platforms.
It’s been said that the new OS design language is taking cues from visionOS, but that isn’t entirely accurate.
Apple
Another thing to keep in mind: Apple is really good at playing the long game. Over the last few years, I’ve had several conversations with Apple executives who have explained that when Apple designs a new processor, it knows exactly what devices will be using that processor. This is one of the advantages of the Apple silicon approach. Well, Apple’s new design language presumably has been designed with some understanding about where the company is going with its product designs in the future, not just in 2025. If Apple is planning a folding iPhone, this design should have factored that in. If there are touchscreen Macs on the horizon, this would be the time for the design to take that into account.
Finally, a lot of pundits will say that Apple’s new design, when it’s revealed in June, is a smokescreen designed to distract people from Apple’s rough time with Apple Intelligence. This argument is about as convincing as saying that Apple changed the iPhone’s design a month before shipping, or altered the design of a chip six months before it arrived. So much of what Apple does takes years of pre-planning to accomplish, and that includes this new design.
But that said, if the new design gets people talking about something other than the broken promises of Apple Intelligence, Apple’s not going to complain one bit.
