The Form Factor Is Eyewear

The industry has settled on a story about what comes after the phone. It has two chapters, and they are both about where to carry the thing.

At Connect in September, Meta announced that the Meta Ray-Ban Display, the glasses with a small screen in one lens, would go on sale in the UK and Canada on 23 September, with France, Italy and Germany following on 13 October. The UK price is £749. The same keynote introduced Muse, an agent that acts on what the wearer is looking at, and promised more than a hundred glasses styles by the end of the year. A Road to VR report adds a voice-driven hologram calling feature, in early access on WhatsApp later this autumn.

The other chapter is a rumour with a very good tailor. According to Bloomberg’s Mark Gurman, as summarised here, OpenAI still expects to show its first hardware this year and ship it in 2027. It is a screen-free companion that lives in the pocket, the product of the roughly $6.5bn io acquisition, and Jony Ive has been calling it a third core device. Apple’s lawyers are in the story too, with a trade-secret suit that names hundreds of former Apple staff who moved across.

Two bets, one assumption

Meta bets on the face. OpenAI bets on the pocket. Both assume that the next layer of computing is a device that a person carries, wears and authorises, and both are probably right about the market.

Neither is ambient computing. A pair of glasses with a display is a screen that cannot be put down. A pocket companion with no screen is a phone that has lost its glass and kept its attention economy. Weiser’s argument in 1991 was that the technology should recede into the world, and you cannot recede while you are strapped to a head.

The room has no vote

Here is the design problem the keynote did not touch. The announcement post I read covers price, styles and the agent, and says nothing about the people standing in front of the wearer. A device that acts on what its wearer is looking at is, by construction, acting on what is in the room: the colleague, the child, the stranger at the next table. They agreed to nothing.

This is the old question that Soft Systems, Hard Questions tries to raise, now with a retail price. Whose consent does an ambient system need? The wearer’s, obviously. The room’s, if the room is where the sensing happens. A consent model built for one person with one device does not stretch to cover that.

What we would prototype

Our view, and it is a view: the interesting work is not a better pair of glasses. It is the signal a pair of glasses gives to the people near it. A bystander needs to know when the thing is looking, what it is allowed to do with that, and how to say no without making a scene. That is a design brief for objects and light, not a settings screen, and it is the kind of thing the Pip was built to argue about.

If the post-screen era arrives on faces, then the face needs manners. Someone should design them before it ships a hundred styles.