BY HARRY MCCRACKEN
When I first got excited by Siri, it wasn’t part of Apple, let alone the iPhone. At the time—February 2010—it was just promising a stand-alone iPhone app from a startup that had been spun out of Silicon Valley R&D icon SRI, drawing on its research collaboration with the U.S. Department of Defense’s DARPA lab. Almost three months elapsed before Apple acquired Siri, and another year and a half until the company built it into iOS, starting with a beta version on the iPhone 4s.
I was excited about that, too, calling it “breathtaking for a beta” and adding, ”If voice-activated assistants are all around us in five or ten years, we’ll look back and say it all started here.” They were, and we did.
But long before Siri celebrated its fifth birthday, its reputation foundered. Even early on, much analysis of the voice assistant deemed it a disappointment, often expressing the hope that Apple would eventually give it a transformative upgrade. Over time, oceans of wordage were dedicated to the topic.
More than 15 years after I first gave Siri a try, it’s still waiting for its big, game-changing update. In this case, the update in question is the more “natural, relevant, personal” version that Apple previewed last June as part of “Apple Intelligence” during its WWDC keynote. It still hasn’t shipped. And on Friday, the company announced that work on the update was taking “longer than we thought” and the release wouldn’t happen until sometime in “the coming year.” The most logical guess: It will be rolled into iOS 19 and MacOS 16, which should ship this fall.
Over at Daring Fireball, John Gruber has a long and acidic account of the delay and its implications. In brief: Apple seems to have repeatedly shown off stuff so far from completion that it wasn’t even ready for live demos. Eventually, the company concluded that it was in over its head—for reasons it hasn’t explained, and won’t—and pushed the release off to some unspecified date.
There’s a name for products like that: vaporware. The tech industry is rife with examples. Apple, in its modern history, has been atypically disciplined about avoiding them—which makes this incident only more striking.
Now, it’s easy to understand how Apple bit off a more ambitious upgrade than it could chew its way through as quickly as expected. The new Siri is designed to respond to free-form requests such as “Send Erica the photos from Saturday’s barbecue,” a big leap from the assistant’s history of only understanding a limited set of instructions expressed in a precise way. Along with requiring greater language skills, the new Siri will sift through your email, calendar, contacts, notes, photos, and other information stored on your device in ways that haven’t been done before.
All this is the most interesting use of AI that Apple has announced, and also the most ambitious. It’s literally something only Apple could do. No other company has enough access to iOS to dream of building it—though as my colleague Jared Newman explained last June, it also held the potential to make third-party apps way more Siri-friendly than in the past.
But for all of the new Siri’s potential, it also feels like something that Apple was scrambling to release as proof it isn’t behind in AI. The generative AI boom unleashed more than two years ago by ChatGPT has left the company in an unusually reactive mode, as it plays catch-up in areas such as image generation. A much better Siri might have taken Apple far beyond me-too territory. It still could. With the delay, however, the company only looks more like it doesn’t yet have a handle on AI and how to make the most of it in Apple products.
I can’t help but think, though, that Apple’s failure to get the new Siri out the door isn’t just about the challenge of doing AI in a way that’s useful, reliable, and safe. It’s part of the much larger, longer story of Siri being full of promise and only sporadically living up to it. I see two other specific factors at play.
Factor one: Apple may be overwhelmed by the sheer quantity of software updates it tries to pump out each year. In the pre-iPhone era, the company had only one operating system to wrangle—MacOS, then known as OS X—and didn’t attempt to update it on a set timetable. Now it’s put itself on a yearly schedule and must juggle upgrades for MacOS, iOS, iPadOS, WatchOS, tvOS, and VisionOS. Of course that’s tough. It’s no shocker that certain elements might suffer from insufficient resource allocation: More often than not, for instance, my beloved iPad feels neglected from a software standpoint. And Apple TV—theoretically a core Apple product in the streaming age—remains a little-changing hobby.
Artificial intelligence must be the furthest thing from an area Apple feels it can safely deprioritize. But it’s also among the most demanding. The company may simply have had too many things going on at once to adequately focus on Siri, even after bragging about the new version during its WWDC and iPhone 16 keynotes.
Factor two: On some level I don’t quite understand, Apple may never quite have emotionally bonded with Siri. How else to explain the company’s failure to do all that much with it over all these years?
Acquiring Siri in 2010 was prescient. So was building it into the iPhone. If Apple had pushed ahead with the feature as ambitiously as possible, iOS might look quite different today. Apple might even have a reputation for being ahead in AI. But maybe Siri simply took it out of its comfort zone of polished visuals, touch input, consistent experiences, and the other elements that made the iPhone such a landmark.
Declaring that Siri was headed for “a new era,” showing it off in splashy canned presentations, and then kicking the can down the road is one of the more embarrassing predicaments Apple has created for itself in recent years. Despite that, I think the delay was sensible. It’s more important that Siri be great than that it arrive on time. And having waited for it to be great since 2011, we can surely wait a little longer.
