# Was Siri just thirteen years too early?

> Source: <https://asymco.com/2026/06/21/was-siri-just-thirteen-years-too-early/>
> Published: 2026-06-21 13:40:35+00:00

*An Office Hours question asked by Bob V., June 9, 2026.*

**Q: When the iPhone 4S came out, the rumor was that Siri was its own operating system running alongside iOS. Watching WWDC, that intent-based OS finally seems to be arriving. Was Siri just thirteen years too early?**

It would seem that there’s nothing really new here but a quality and performance increase–admittedly a quantum one. Siri’s original dream, back when it was a research project at Stanford, was very much along these lines (as was Alexa, et. al.). I started my own career in information retrieval, a natural-language interface to query computers before we even had search. The idea was that you shouldn’t need to know how to program, or how to use a database query; you just ask for what you need and get answers sorted by relevance. All of that existed in the 1980s.

So what’s new isn’t the chatbot. You can build a neat chatbot that fools people into thinking they’re talking to a human. What we started to see at WWDC is something else: a commitment to re-architect the entire guts of the OS, across all of Apple’s operating systems in order to surface this experience.

Subscribe to **Asymco One** for full access to our live Q&As.

But I came out of it asking: show me the money. None of AI is interesting unless we see economic activity, because if there’s no benefit, people treat it like another unrequested, unloved, unused feature. That’s the bane of product development. An unused feature is a symptom of over-service — you deliver things people don’t take up, and therefore won’t pay for. Christensen put it well: a commodity is a product where an improvement in performance does not allow you to increase its price. Iowa corn might be better than Kansas corn, but on the Chicago exchange it’s all just corn. All commodities have the same performance; a pork belly, a barrel of oil, a kWh, a bushel of corn. You don’t want that to happen to your product or service. Improvements must be both used and valued by the customer.

Apple is a consumer company with about 1.4 billion users. If they treat AI like Memoji — cute, but not life-changing — there’s no economic impact. I do however think Apple Intelligence is coming together into something people will use daily, and then we’ll see how much they’ll pay for this extra power.

Right now the simple option is to pay for new devices to run it. Eventually, however, it will get repackaged into Services–probably as a subscription to more access, and that’s where the recurring revenue lies. iCloud already bundles a compute allocation; expect those tiers to change, and expect more price elasticity to enter the picture.

*Editor’s Note: This was one of the questions asked by participants in Asymco’s June 2026 Office Hours live Q&A session, open to Asymco One subscribers.*

Subscribe to [ Asymco One](https://asymco.com/one) for more in-depth analysis and commentary like this.
