Dear Reader,

Fires were burning in Cupertino this March as news of an internal meeting revealed consternation within Apple over the state of development for the Siri voice assistant. The situation was described as "dire" in the headlines, although the meeting appears to have been more like a combination of a pep talk and an expectation-management exercise. Unfortunately, whoever leaked the news of the meeting leaked it to Bloomberg, who placed the article behind a paywall [1], so most of the descriptions are from third parties [2], but it appears that the "incredibly impressive" new features targeted for release with iOS 19 "only work properly up to two thirds to 80 percent of the time" [3].

It is embarrassing, of course, for this kind of thing to leak, but I can't believe it's all that rare for companies to get behind on where they thought they would be and have to have an all-hands meeting over it. Especially now, in this era of artificial intelligence, when software managers are forced to use conventional time estimation techniques to guess how long it will take to develop very unconventional new features.

Many of these new features appear to bring increased sophistication to the sorts of tasks you can get Siri to do based on voice commands – similar in spirit to the scene at the beginning of Blade Runner when Harrison Ford's character zooms in on an image by talking through it with his TV. The new Siri is also supposed to be able to infer more from the user's context, including taking actions based on the contents of the screen. The biggest problem isn't that these features are late – it's that they have already been announced (or teased, anyway) in early advertising for the iPhone 16, and it doesn't look good to boldly go where no one has gone before and then have to take a sheepish step back. Still, it seems the company will probably survive. I know this won't sit well with the curators of Apple's futuristic self-image, but I'm not sure people even buy iPhones anymore for their cutting-edge technology – they buy them because they are reliable and steady, and this argues for slowing down and making sure you get it right.

As for these revolutionary new "convenience" features making their way into the next generation of phones, not wishing to sound like a hermit on a hillside (which, admittedly, I am), I can't help but be a bit amused that so-called "visionaries" such as Ray Kurzweil and Elon Musk imagine a future where human brains incorporate elements of silicon technology to form a super-charged hybrid thinking machine, when the real development appears to be more in the direction of people offloading their thinking to external devices.

Despite the futuristic context, the real problem facing Siri and many other AI systems is the same issue that has plagued commercial software development since its inception: marketing overpromising what engineering can deliver in the available time. Apple has lived in this gray zone since its inception. Early Apple employees used to talk about the "reality distortion field" that surrounded Steve Jobs. Actually, the company has been rather good at weathering such crises. In 2012, for instance, Siri faced a class-action suit from disgruntled users who felt the advertising for the iPhone 4S made misleading claims about her intelligence [4]. The suit was dismissed on a technicality, but the impression remained, and Apple eventually persisted and finessed their way through it.

Through the years, the company has stared down such crises by adeptly lowering expectations and simultaneously focusing on development. I'm fairly confident they will find their way through this one in a similar manner. And if they don't, and Siri turns out to be just a little bit lower on the intelligence scale than AI beauty-pageant rival Alexa, something tells me that won't be too much of a problem either. Apple will just give us something else to covet.

Editor in Chief, Joe Casad