What happens when 1.5 billion people suddenly get a good AI assistant in their pocket?
In just a few days time, when Apple holds the keynote for WWDC 2026, we are likely to finally see (and hear) a Siri that doesn’t suck.
While this has been rumoured in the tech press, I don’t think most people are quite ready for how big of a deal this is.
And even if Siri stops sucking, it’s going to take a long time and a lot of advertising dollars to persuade people it’s actually true.*
Apple is a victim of its own doing. They are the boy who cried wolf on AI.
The lowest barrier to entry for AI
For years, the modern wave of AI has been mostly opt-in.
People have had to:
- Download ChatGPT
- Sign up for Claude
- Try Gemini
- Learn new habits
A lot of people have, of course, but many (most?) people still haven’t.
And even for those who have downloaded and used any of those apps, I suspect the majority of iPhone users don’t use AI regularly, or benefit from even a slither of its capabilities.
But everybody knows Siri.
Everybody has a button for Siri.
Everybody has a microphone for Siri.
Everybody has already been trained to understand the concept.
The problem was simply that Siri sucked.
If Siri suddenly becomes genuinely useful, Apple will be transforming an existing habit that they’ve spent years training users to be disappointed by.
It’s potentially enormous, but it’s going to take a while for people to catch on.
Learned helplessness
Siri has created a form of technological learned helplessness.
Users stopped asking difficult questions of Siri. In fact, many users stopped asking anything of Siri.
They learned the answer would usually be disappointment.
If Apple has fixed that, the hardest part won’t be improving Siri, but teaching people to try again.
How is Siri suddenly getting better?
Multiple reports suggest Apple is increasingly turning to Google and its Gemini models to power some of their AI ambitions. Google event tweeted about it:
Joint Statement: Apple and Google have entered into a multi-year collaboration under which the next generation of Apple Foundation Models will be based on Google's Gemini models and cloud technology. These models will help power future Apple Intelligence features, including a…
— News from Google (@NewsFromGoogle) January 12, 2026
You can look at Google’s latest announcements to get a sense of what’s coming to the iPhone. If you squint, repeat the announcements in a Tim Cook voice, and apply some Liquid Glass to the UIs, you’re on the right track.
Apple is essentially forming their own variant of a multitude of Gemini features and models.
If Apple doesn’t screw this up, then WWDC is sure to be a demonstration of great AI functionality actually coming to iOS and Apple’s other platforms.
You can also look at the recent accessibility features Apple announced for a flavour of what they’re capable of now thanks to their latest efforts.
This time, it’s real people.
How do we not end up in the same position 2 years ago?
In a seemingly rushed segment of the WWDC 2024 keynote, Apple introduced Apple Intelligence, and showed off several AI features that got people excited.
See this from Apple’s press release on 10th June 2024:
Siri will be able to deliver intelligence that’s tailored to the user and their on-device information. For example, a user can say, “Play that podcast that Jamie recommended,” and Siri will locate and play the episode, without the user having to remember whether it was mentioned in a text or an email. Or they could ask, “When is Mum’s flight landing?” and Siri will find the flight details and cross-reference them with real-time flight tracking to give an arrival time.
And … **checks notes** Apple still hasn’t shipped these features.
It was a rare misstep for Apple, making it a fascinating moment to discuss and decipher. John Gruber summarised the whole situation effectively in his damning but fair piece Something is rotten in the state of Cupertino.
This time, I believe Apple has learned its lessons. They won’t show a single feature they can’t demonstrate live in front of people. There’s just no way they can burn that trust again. The only way to put the record straight now is to show and tell with a live, working experience, end to end.
iOS 27: the one where Siri finally stopped sucking
While the challenge for Apple for years has been to build a fundamentally better Siri, I feel remarkably confident that this time, they might well have done it.
But this, of course, is no longer their real challenge.
The challenge now is to overcome more than a decade of disappointment.
Because if Siri really is good now, most people won’t believe it. At least not at first,
They’ll have to discover it for themselves, one successful interaction at a time.
It’s going to be a fascinating year ahead to observe Apple. They’re not just launching a new AI assistant, they’re attempting their biggest reputation turnaround in modern times.
*Apple is no stranger to spending advertising money on AI features. It might be easier to promote a better Siri that actually works if they didn’t run a whole campaign around AI features that don’t even exist. I guess that’s what one might call karma.