Jacqueline Roy, an Apple spokeswoman, in a written statement to John Gruber at Daring Fireball:

Siri helps our users find what they need and get things done quickly, and in just the past six months, we’ve made Siri more conversational, introduced new features like type to Siri and product knowledge, and added an integration with ChatGPT. We’ve also been working on a more personalized Siri, giving it more awareness of your personal context, as well as the ability to take action for you within and across your apps. It’s going to take us longer than we thought to deliver on these features and we anticipate rolling them out in the coming year.

This “more personalized Siri” featuring Apple’s personal context artificial intelligence feature was demonstrated at the Worldwide Developers Conference last year as part of iOS 18 and was supposed to be out by the spring. Apple never acknowledges future products — for all intents and purposes, this is supposed to be a current feature of iOS just coming in a future update. But reading the tea leaves leads me to believe that the new Siri won’t ship until the fall, perhaps as part of iOS 19, which is rumored to feature no substantial improvements to Apple Intelligence. This year’s WWDC invites should be out in a few weeks, and Apple’s development teams are all hard at work on putting the final touches on the next operating systems scheduled to go into beta in June. There’s just no time to ship these 18.x features by the end of the month.

But why? Apple’s demonstration last June was supposedly recorded live, so the features must’ve been partially developed by then — or at least, that’s what I thought. Most likely, WWDC was the work of Apple’s marketing department with no oversight from engineering, similar to the AirPower wireless charging mat announced alongside the iPhone X in 2017 that died because it was impossible. Phil Schiller, Apple’s then-marketing chief, proudly proclaimed that his teams “know how to do this” on that September morning, just like Craig Federighi, Apple’s software vice president, said about Apple Intelligence. Neither of them seemed to be correct in their assertions. Here’s Mark Gurman reporting for Bloomberg on the internal shenanigans at Apple:

Since then, Apple engineers have been racing to fix a rash of bugs in the project. The work has been unsuccessful, according to people involved in the efforts, and they now believe the features won’t be released until next year at the earliest.

In the lead-up to the latest delay, software chief Craig Federighi and other executives voiced strong concerns internally that the features didn’t work properly — or as advertised — in their personal testing, said the people, who asked not to be identified discussing internal matters…

Some within Apple’s AI division believe that work on the features could be scrapped altogether, and that Apple may have to rebuild the functions from scratch. The capabilities would then be delayed until a next-generation Siri that Apple hopes to begin rolling out in 2026.

If Gurman’s reporting is to be believed, there’s no functioning version of the “new” Siri at Apple Park. Marketing seems to have caught on; an advertisement from last fall showing the actress Bella Ramsey asking questions to the more personal version of Siri was deleted this week upon the news of the statement. The fancy rainbow glow at Apple’s Fifth Avenue store — “when I say A, you say I” — the “Hello, Apple Intelligence” slogan on all of last year’s iPhone models, and the countless occurrences of billboards in subway stations and city streets around the country were all for nothing. They advertised nothing. Those features never existed beyond a nonsense presentation cobbled together in a few months.

I gave Humane a hard time for shipping the Ai Pin in a state where it was nothing more than overpriced vaporware. The Ai Pin does nothing it was supposed to and I rightfully flamed Humane for its marketing lies and subsequent pump-and-dump scam. I’m applying my standards evenly: Apple sought to capitalize on the AI stock market gold rush last year, put together a fancy demonstration for eagle-eyed WWDC viewers, and never shipped the feature. The only difference between the two companies is that I still presume Apple intended to release them in the spring, whereas Humane’s primary motive was to find the first idiot to buy the company for an obscene amount of money. Either way, the outcomes were the same.

This is sheer, unbridled incompetence, and there’s no other way to put it. John Gianandrea — Apple’s machine learning chief who has accomplished almost nothing during his time there — Federighi, and Tim Cook, the company’s uncharismatic, slow-as-a-snail chief executive, all need to huddle and figure out how to address their collective lack of meaningful leadership skills. Otherwise, they should all be fired. I’m not saying this as a fluke — if Scott Forstall, Apple’s previous software chief pre-Federighi, was forced out of Apple due to the cataclysmic Apple Maps failure, Federighi and Gianandrea should be out of the door by next week. The only difference was back then, in 2012, Apple still had a culture dominated by the Steve Jobs school of thought. People who didn’t do good work were sacked with no remorse. Apple under Cook is ruled by a hierarchy of convoluted politics, whereas Jobs governed like a monarch.

Federighi, Gianandrea, and Cook’s roles in the Apple Intelligence drama have significantly undercut Apple’s leadership in software. That politics is, in my eyes as an outside spectator, the real reason why Apple has no AI strategy. There’s only one solution I can think of: Gut the entire department and buy another company that knows how to make large language models. Of the ones out there — or the ones for sale, anyway — Anthropic jives the best with Apple’s ethos of privacy and safety at the heart of every core innovation. Claude is designed to be safe — so safe, in fact, that it doesn’t even search the web yet. I’ve written this before, but the problem doesn’t lie in the low-level programmers who write code in Xcode or fiddle with models, but rather the people with oversight tasked with managing the direction of the company. Anthropic’s direction is very similar to Apple’s, if it had a direction at all: build artificial general intelligence that’s safe and benefits humanity. Apple just needs to adopt the fast-paced nature of a Silicon Valley start-up. Cook’s politicking both inside and out needs to go.

Apple doesn’t have a money or a people problem. Gianandrea worked for years at Google, developing some of the best machine learning in decades. Federighi’s teams are efficient and produce beautiful and intuitive software used and relied upon by billions daily. Cook’s leadership has turned Apple into the most valuable company in the world. The engineers who do the scut work make products worth praising day in and day out. They’re some of the most talented, smart, and creative people in software development. On paper, Anthropic pales in comparison to Apple’s efficiency and knack for great hires. But Apple needs a direction after playing internal politics for far too many years. There’s too much bureaucracy in that company. Every report from Gurman leaves me feeling like the low-level programmers at Apple are trying to send a message that the C-suite just isn’t interested in doing the work. They’re not worried. They’re never worried. They’ve become too comfortable in their company’s place as top dog. They need a fire under their seat, and the people at Anthropic have exactly that.