Welcome back to the Even Insider.
Most hardware ships finished. What a device can do on the day you open the box is usually everything it will ever do - a feature list frozen at the factory. The exceptions are rare, and they define eras. A phone is not valuable because of what its maker built into it; it is valuable because of what everyone else built for it. The moment a device opens itself to the people who use it, it stops being a product and becomes a place.
In March 2026, Even G2 joined the exceptions. We opened Even Hub inside the Even Realities app - a software ecosystem for the glasses you already own, open to anyone who wants to build for it. This entry is about why we did it, and about what changes for the thing on your face.
The Ceiling Problem: Why a finished device can only run out.
Every fixed-function device carries an invisible ceiling, and the ceiling is the maker's imagination. If they did not picture your situation, your situation does not exist for the device. A team, however good, is finite. A roadmap, however ambitious, is a list, and a list has an end.
We felt that ceiling from the inside, before anyone outside did. The requests that reached us were never random - they clustered into whole worlds of need we could see clearly and could not possibly all build ourselves. The traveler, the student, the person who reads on trains, and the person who runs before dawn each wanted something specific and reasonable. The uncomfortable arithmetic of a fixed device is that the features worth the most to a person are usually the ones that cost the most to build. A single team can fund a handful of them a year. Not the fortieth. Not the four-hundredth. The ceiling is not set by ambition. It is set by arithmetic.
There is also a slower cost. Before a platform, a need and the person who could meet it have no way to find each other except the expensive machinery of a contract - and most good ideas die inside that machinery, not because they were wrong, but because they were never the top of anyone's list.
Opening the platform changes the arithmetic and dismantles the machinery at the same time. When anyone can build for the glasses, the ceiling stops being a function of our roadmap and becomes a function of every builder who shows up. A company that builds everything itself is a pipeline: it can only ever ship what it can staff. A platform is a marketplace of effort, where a need on one side meets the time and the skill to solve it on the other - without a meeting, without a contract, without a quarter of planning in between.
Opening the glasses does not mean we step back from them. The two paths run in parallel, by design. Inside Even, we keep polishing the core conversational experience - the dialogue between you, the glasses, and the intelligence behind it - to the depth only the maker of the hardware can reach. Even Hub runs in the other direction: outward, into the width of the ecosystem, the diversity of use cases no single roadmap could cover. Depth from us. Breadth from everyone who shows up. Neither path waits for the other.
This is where the glasses stop being a product and become a place. A pipeline device depreciates, because its abilities were frozen at the factory. A platform device appreciates, because every app that ships for it is a capability the hardware did not have yesterday and has today. The glasses on your face can be worth more this month than they were last month. The reason is other people.
A Corner of Your Vision: How the platform actually works.
An app on Even Hub borrows a corner of your vision and a few gestures of your hand, and gives them back the moment you are done.
Start with where the work happens. An app you install from Even Hub does not run on the glasses. The glasses have no room for it, and it does not need to. The app runs where the power already is - on your phone, inside the Even Realities app. Even G2 is the surface it draws on. Even R1 is the hand that steers it. Three devices, three jobs: the phone thinks, the glasses show, the ring decides.
There is a hardware reason the app lives on the phone and not on the glasses, and it is the same reason Even G2 can be worn all day and forgotten. Compute is heat, weight, and battery. A thing that sits on your nose can spare none of the three. So the glasses stay deliberately thin - a display and a set of sensors. The thinking is done by the computer already in your pocket, which has power to burn. The division is not a compromise we settled for. It is what lets the glasses stay glasses.
An Even Hub app is a guest in the house, and a guest does not get keys. It runs sandboxed; it does not inherit the core app's permissions by default. If it wants the microphone, it must ask for the microphone - explicitly, in its code, in the open - and you must grant it before the app can run. The same holds for network access. Nothing reaches your data because it happened to be running nearby.
Then there is the canvas, and this is where a developer meets the real Even G2. The display is 576 by 288 pixels, one color, monochrome green. That is the entire stage. Input is exactly four gestures on the Even R1 - tap, double tap, swipe, long press - and every app on the platform shares them. That last detail matters more than it sounds. Because the gesture grammar is fixed across the platform, muscle memory transfers. You already know how to drive an app you have never opened, because it steers like every other one.
To a builder arriving from a world of infinite canvases, this can look like a cage. It is closer to a gift. A large screen invites clutter; it lets you postpone the hard question of what matters most, forever. A 576-by-288 canvas refuses the postponement - it forces the decision on the first day. Four gestures cannot be bent into a hundred obscure combinations, so nobody ships the combination nobody can remember. The narrowness of the SDK does not stop developers from building good apps. It stops them from building bad ones. We did not stumble onto these limits and apologize for them. We drew them on purpose, because they do a share of the design work on every builder's behalf before a single line is written.
The Cost of Entry: What it takes to build.
A platform lives or dies on a single number that no one outside it ever sees: the distance between having an idea and watching it run. Every step across that gap - every account, every dependency, every device you must own before you can begin - is a place where a builder quietly gives up. So we spent our effort collapsing the gap, on purpose, to as close to zero as we could get it.
The SDK is TypeScript - @evenrealities/even_hub_sdk - a language most developers already know, not a private dialect they would have to learn before they could even start. That one choice erases weeks.
The simulator erases a device. There is a desktop simulator that renders your app exactly as the glasses would, on your laptop, so you can build for Even G2 without owning Even G2. The barrier to your first app is not a purchase; it is a terminal window. You can write it, run it, and watch it paint the monochrome canvas.
The package is a single file. A finished app compiles to one .ehpk - one file to build, one file to hand a friend, one file to submit through the developer portal. There is no sprawling bundle to assemble, no ceremony to the handoff. The unit of a developer's work is small enough to attach to a message.
None of this is generosity. It is the answer to the oldest problem a platform has. Every platform begins cold: nobody wants to browse a place with no apps, and nobody wants to build for a place with no users. You break that stalemate exactly one way - by making the cost of building so low that people build before there is an obvious reason to. Out of curiosity. On a weekend. Because the barrier was lower than the itch. Once the first apps exist, the loop turns on its own: apps bring users, users bring builders, builders bring apps. We chose quality over volume, because a loop that turns is worth more than a shelf that is merely full.
You can already see the shape of what that produced. In the first months of the open platform, the earliest builders did not wait for permission to be interesting. They shipped things to read - long text, paced and legible, on a surface that sits quietly at the edge of sight. They shipped things that keep time - timers and counts and intervals for people whose hands are busy. They built small windows of ambient information that live at the border of attention and ask for none of it. They built things purely to play, because a new canvas is an invitation and someone always accepts it. Not one of those was on our roadmap. That is the whole point.
What's Next?
The glasses we shipped in April were not finished. We want to be exact about that, because it is easy to hear it as an apology when it is the opposite of one. A finished device is a device that has stopped becoming more useful - its peak already behind it, on the invoice. We shipped Even G2 deliberately unfinished, with the door left open, so that its best day would always be somewhere ahead of it. Depreciation is what happens to hardware that was completed at the factory. Appreciation is what happens to hardware that a community is still building.
So the invitation splits, depending on which side of the glasses you stand on. If you wear Even G2, open Even Hub in the Even Realities app and look at what has been built for the thing you already own - some of it by people solving a problem you happen to share. If you build, go to the developer portal, open the simulator, and ship your first .ehpk before you have talked yourself out of it.
Next month we turn the argument inside out. We have spent this entry on everything the glasses can now become. The next one is about everything they refuse to be. A platform is defined as much by what it declines to do as by what it allows - and restraint, it turns out, is also a design.
