Why the Next iPhone Won’t Be an iPhone

Ive, Altman, and the quiet category shift Apple isn’t ready for.

Photo by Gilles Lambert on Unsplash

Every six months, someone declares the smartphone dead. The Humane AI Pin was going to kill it. The Rabbit R1 was going to kill it. Meta’s smart glasses were going to kill it. None of them did. The phone is still in your pocket, and you’ve probably looked at it twice while reading this paragraph.

So when Jony Ive and Sam Altman started building a device together at io — now owned by OpenAI for $6.5 billion — most of the coverage framed it as the next attempt at a smartphone killer.

That framing is wrong. And in the gap between the wrong framing and the right one sits the answer to a much sharper question: what is the next iPhone going to be, if not an iPhone?

The smartphone is not going to be replaced. It is going to be surrounded. And the device that takes the central spot — the new gravitational center of personal computing — won’t look anything like the rectangle in your pocket.

The pattern everyone misses

Look at what every successful new computing category has actually done. The smartphone didn’t replace the desktop PC. The desktop PC sat on your desk, the smartphone sat in your pocket, and over time the PC quietly shifted from “primary computer” to “specialized tool for when you really need to focus and type.” The PC didn’t die. It became less central.

The same thing happened with the watch. The Apple Watch didn’t replace the iPhone. It surrounded it — handled the quick glances, the workouts, the notifications, the things your phone was overqualified for. And the AirPods didn’t replace the phone either. They surrounded it from the audio side.

This is the actual pattern. New categories don’t kill old ones. They take specific use cases away, one at a time, until the old category is reduced to a smaller slice of your day.

The phone has had an enormous slice for fifteen years because no other device could match its versatility. That’s about to change — not because someone builds a better phone, but because someone builds a device that takes one slice and does it dramatically better.

What Ive and Altman are actually doing

The leaks are now consistent enough to triangulate. Ming-Chi Kuo says the device will be slightly larger than the Humane AI Pin but as compact and elegant as an iPod Shuffle, with no screen. Altman has called it “peaceful” and said users will be shocked at how simple it is. The release window is the second half of 2026.

But the most important detail is one most coverage skipped. In an internal meeting, Altman and Ive told the OpenAI team that the device is designed to be the user’s third device — something you put on your desk alongside the laptop and phone you already own. Altman even said it would be a “family of devices,” not a single product.

This is a remarkably honest positioning. They are not pretending to replace anything. They are claiming a third spot, a new slice, a new context.

That third spot has a specific shape: voice as the primary interface, context-awareness as the differentiator, and ambient presence as the form factor. It is what the smartphone is bad at — because the smartphone is a screen-first device that has to be picked up to be used.

Why voice finally works

Voice as a primary interface has failed for fifteen years. Siri, Alexa, Google Assistant — they were all worse than just opening the phone. The reason wasn’t the microphone or the speaker. The reason was the model. Voice assistants were brittle, literal, and limited to a small set of commands. Talking to them was slower than tapping.

That has changed in the last two years. The new generation of audio models can hold an actual conversation. They handle interruptions, understand context, recover from misunderstandings, and respond with emotional nuance. According to The Information, OpenAI is building a new audio model architecture for exactly this device — one that can speak at the same time as a human and handle interruptions like a real conversation partner.

When voice becomes genuinely better than tapping for a use case, that use case migrates. And there are a lot of use cases — capturing a thought, asking a quick question, dictating a draft, planning a day — where speaking has always been faster than typing. We just didn’t have a system that could keep up.

The screen will not die — it will become invisible

Here’s where I expect to disagree with most futurists. The screen is not going away. People still want to read, watch films, look at photos, see a map. What is going away is the screen as the primary interface to a computer.

The screen has existed for forty years for one reason: computers couldn’t understand intent directly. So you had to translate your intent into a specific application — open Maps, open Spotify, open WhatsApp — and use a UI as the bridge between what you wanted and what the machine could do. The app was a workaround for the limits of the machine.

When the machine can understand intent directly, the app stops being necessary. You don’t need to “open Maps and search for a restaurant and call to reserve and add to calendar.” You say “book us somewhere nice on Friday near the office, tell Anna, put it in my calendar.” One intention. One sentence. Done.

The screen becomes a verification surface and a consumption surface. You glance at it to confirm what the agent did. You watch a film on it. You don’t drive your computing through it anymore.

Apple’s quiet trap

The most overlooked move in this whole story is what Apple is doing right now. Bloomberg reports that iOS 27, expected to be announced at WWDC on June 8, 2026, will introduce an Extensions system. It will let users choose between ChatGPT, Claude, Gemini, and other AI providers as the engine behind Siri. Apple has effectively given up on winning the model layer and is positioning itself as the router for AI demand — collecting App Store revenue from every AI subscription that flows through its devices.

This is brilliant in the short term. Apple makes money without having to win. The iPhone stays central. The user gets choice. Everyone wins.

It is also a textbook innovator’s dilemma. As long as Apple is making money from being the router, it has no incentive to cannibalize the iPhone with a fundamentally different device. And while Apple collects rent, someone else gets to build the device that the iPhone is a compromise for.

Because the iPhone is a compromise for ambient, voice-first, context-aware computing. The microphone is in the wrong place. The camera can’t run continuously. The battery is dimensioned for screen use, not for eighteen hours of background listening. The form factor requires you to pick the device up. None of these are fixable through software extensions.

Apple’s bet is that this doesn’t matter for another five years. That bet might be right. It might also be how Apple becomes the BlackBerry of the 2030s.

What the third device actually does for me

Here’s the test I run on every “next big device” claim: would I use it tomorrow, and what would change?

For me — and I suspect for a lot of knowledge workers — the answer is yes, if it solves one specific problem. The problem is that the distance between thinking something and capturing it into a system I can act on is too long. I have an idea while making coffee. I have to remember it, walk to the laptop, open Obsidian, navigate to the right folder, type it out. By the time I do that, three other ideas have evaporated.

A desk-bound, voice-first companion that hears me, understands my context, and writes directly into my system would compress that distance from minutes to seconds. Not because typing is slow. Because the translation overhead between thought and structure is what kills momentum.

That is the third device. Not a phone replacement. Not a magical pendant. A specialized companion for the slice of computing where voice plus context wins, and where the screen is in the way.

The bigger picture

The interesting question is not which company wins the third device. The interesting question is what happens when there is no longer one central device.

For fifteen years, the smartphone has been the gravitational center of personal computing. Everything orbits it. The watch, the AirPods, the laptop — all configured to work with the phone. If the third device succeeds, that gravitational center starts to weaken. The phone becomes one of several specialized devices, none of them dominant.

That is a much bigger shift than “the smartphone gets replaced.” It is the end of a single-device era. It is the beginning of computing as a distributed property of your environment — something Mark Weiser called “ubiquitous computing” thirty years ago, and that finally has the model layer to make it real.

The next ten years won’t produce a smartphone killer. They might produce a smartphone de-centerer. And the people building for that future are not trying to make a better rectangle for your pocket. They are trying to dissolve computing into the rooms you already live in.

The phone stays. It just stops being the answer to every question.