.jpg)
This piece is informed speculation based on reported developments and publicly available intent signals. What follows isn’t a product announcement. It’s a map of what might come next; and what the advertising industry should be thinking about before it does.
GPT can hear, speak, process real-time visuals, and hold context like never before. But even as OpenAI pushes the boundaries of what large models can do, it’s also staging a quieter but potentially more radical shift: building the object that will carry those models into everyday life.
That object is expected to emerge from OpenAI’s acquisition of Jony Ive’s design firm io, in what has been described as a $6.5 billion play to define a new class of AI-native device. Not a smartphone. Not a wearable. Something in between. Something that wants to replace the phone by first making you question why you ever needed one.
The idea of a post-phone world isn’t entirely new; but this version has a few sharper edges. The first io device, according to reliable reporting, won’t be a replacement for your smartphone. Not yet. It will likely pair with it. A minimal, screen-light, voice-heavy object that acts as a companion rather than a conqueror. Its goal isn’t to take over the phone; it’s to take over your relationship to attention. It’s there to train the model on how humans live when they’re not staring at glass.
To understand the stakes, it’s worth remembering how the current paradigm began. Smartphones as we know them were crystallised by the iPhone in 2007. Apple didn’t invent mobile computing, but it reframed it. The BlackBerry, dominant until then, had introduced the concept of a pocket-sized internet-connected device with a full keyboard; beloved by executives, loathed by thumbs (until we learned better).
But Apple removed the keyboard, doubled down on touch, and turned the phone into a lifestyle surface. The screen became the experience. Google, caught flat-footed with its BlackBerry-style Android prototype, pivoted immediately. Android became a touchscreen OS; the app store became the battlefield; and the attention economy took its first real breath.
Now, with io, the same pivot may be coming; only this time, in reverse. If the iPhone made the screen central, the io device is designed to make it optional. If the app store turned functionality into a tap, the assistant will turn it into a presence. And just like the early iPhone didn’t immediately kill the BlackBerry but quietly rewired user expectations, this new device won’t replace smartphones overnight. It will shift where attention lives. And that shift will ripple straight through advertising.
Because if the interface changes, so does everything that rides on it. Advertising is deeply dependent on screens. From banners to video to programmatic display to search, the entire ecosystem is designed around visual interruption and visual measurement. Impressions, reach, CTR, brand lift—it all starts with something being looked at. If that layer dissolves, so does the scaffolding.
A voice-first, context-aware assistant doesn’t scroll. It doesn’t show carousels. It doesn’t nudge you toward a call-to-action button. It offers recommendations. Suggestions. Spoken prompts. And eventually, it just does things. In that world, the ad isn’t a box you see. It’s a moment you hear. Or worse, miss entirely if you’re not part of the assistant’s suggestion logic.
That doesn’t mean advertising disappears. But it does mean it becomes something else. Something less intrusive, more ambient, and much harder to game. A voice nudge that fits the flow of conversation. A well-timed reminder triggered by pattern recognition. Maybe even a branded presence layered into a task the assistant is already performing. In every case, the assistant is the filter. Not the platform. Not the publisher. Not the feed. The assistant.
That’s a power shift; not a product feature.
It also raises questions the industry hasn’t really had to ask before. If the assistant is deciding what a user hears or doesn’t, what a user sees or skips, how do you advertise to the assistant? Is it earned trust? Paid placement? Data proximity? Brand memorability? And how do you measure success when the ad can’t be seen, only felt?
The existing ecosystem doesn’t have answers. It wasn’t built for this. Its tools are tuned for attention metrics, not memory graphs. But the post-phone world doesn’t care about your impressions. It cares about whether the assistant remembers your brand when the user asks; or before they do.
And that changes the role of creatives. It means writing for tone and cadence, not colour and layout. It means building interactions that feel like utility, not interruption. And it means thinking about brand presence as something that trains a model, not just persuades a user.
The io device, if it lands as projected, will not flip the table overnight. It’s not trying to. The first version is more testbed than revolution. A bridge. It will learn how users respond to voice-only interactions, haptic cues, and ambient nudges. It will study how trust is built in an interface with no screen. It will experiment with how brands might exist without being seen. And most importantly, it will keep OpenAI in control of the interface layer that every other platform currently fights to own.
This might sound abstract now. But consider how quickly the last UX paradigm shifted. Mobile advertising didn’t exist at scale until the iPhone put the internet in people’s hands. Social advertising didn’t explode until feeds became sticky enough to dominate attention. Voice-first assistants haven’t yet created a stable ad format, but that’s because they haven’t created a stable user habit. io might change that. Not because it invents something new, but because it makes something ambient.
The opportunity for advertisers is to get ahead of that moment. To start testing formats that don’t need to be seen to be understood. To explore brand tone not as a line in a deck, but as a sensory layer in a conversation. To stop optimising for views, and start designing for recall.
There are real concerns to flag. If OpenAI’s assistant becomes the default interface, who governs what’s surfaced and what’s not? How is bias handled? Who gets access to prompt real estate? These are thorny, unresolved questions; and the fact that no one in advertising is currently asking them should be cause for concern. When the interface collapses into the model, distribution becomes invisible. And invisible systems are notoriously difficult to audit.
But those debates are likely a step away. What matters now is watching the bridge device carefully. io won’t be sold as a media product. But it will quietly become the new place attention flows through. And the brands that adapt earliest will be the ones the model learns to keep close.
No one’s declared the end of the phone. But someone is building its successor. And if advertising doesn’t prepare for a world where the interface listens instead of glows, it may find itself optimised for an audience that no longer looks.
.jpg)
