Or: the phone I didn't want to buy my son
My son had been asking for a phone for almost a year, and I kept hearing the wrong thing. He meant: the device that goes with me. I heard: the device that would eat his attention alive.
It was never screens in general. I am the last person who could make that argument. My worry was specific. My son can read for three, four, five hours straight. He devours books. He disappears into math the same way. And I know what sits behind the apps on a modern phone: teams of smart people whose work is measured by how often he comes back and how long he stays. He would be carrying their best efforts in his pocket, all day, every day. I did not know if he could resist that, because almost no adult I know can.
He already had an iPad. At home, on Wi-Fi, that had been enough for a while. But an iPad is still a place you go to. A phone goes with you. It sits in your pocket at school, on the walk home, at soccer, in the small gaps where thirteen-year-olds now make plans and test friendships. It is not just a smaller computer. It is a way of being reachable.
His friends had one group chat that never really ended. When he was not in it, he was not exactly cut off. He just came in late. The plans had formed. The joke had moved. Someone had already explained the thing that made the next thing funny. He did not make a speech about being left out. He checked later, asked what had happened, and tried to catch up to a conversation that had already moved on.
My wife and I had been circling this for a year. The line we had drawn was the end of seventh grade. Now it was weeks away.
I wanted to give him an Android phone. I had reasons, and I'll get to them. What I found instead is that I more or less couldn't. Not without making him the kid who turns the whole chat green. The thread changes. Some features disappear or work differently. Everyone can see whose arrival did it. I did not want every message he sent to announce that he was the reason the room had changed.
The path of least resistance was sitting right there: buy the iPhone, close the question, move on. I nearly took it. Then I caught myself, because of who I am and what I've spent my life doing. I used to make phones.
Back in 2007, before Android reached the market, I worked on Openmoko. We were trying to build something radical for the time: a phone you could open all the way down. No black boxes. A device you could understand, rebuild, and trust from the ground up.
We thought the answer was openness all the way down. It still wasn't enough. Before the iPhone, and for years after, there was almost no way to scale a phone without going through the carriers. They did not just control access to the network. They controlled the interface, the software, the logistics, the sales channel, and the relationship with the customer. If you wanted to reach ordinary people, you entered their system.
The radio created another set of walls. We were working on what the industry called 2.5G. For a company our size, moving to 3G would have meant spending many millions before we had built a single phone: new radio technology, test labs, carrier certification, and licenses for standards-essential patents held by Qualcomm and others. The network was gated. The radio was gated. The patents were gated. The route to the customer was gated.
Openmoko was our attempt to punch through all of it at once. In theory, everything was possible. In practice, every road was owned by someone who had not invited us.
Years later, working with HTC, I met a more polite version of the same trap. We were trying to build a blockchain phone. The premise was not simply stronger security. If the phone was going to hold keys and sign transactions, the person holding it had to be the final authority. Not Google. Not the carrier. Not even HTC. The point was not to put a more benevolent company in the middle. The point was to remove the middle.
Very quickly, we ran into the same binary. Android was open on paper, but the parts people expected—the Play Store, maps, push notifications, Google Play services—came with Google's compatibility rules. Step outside those rules far enough to give the individual control over the security model, and you risked stepping outside the Android ecosystem that made the phone useful. For HTC, Android was not a side project. It was the business. Asking the company to make that break meant asking it to put the rest of its phone business at risk.
Android still left more room to maneuver than iOS. You could sideload software and replace more pieces. But it was the least bad option, not an open one. The normal path still ran through an app store, a platform account, and rules set somewhere else.
That is not the web that made Google possible. On the web, anyone could publish, link, and build without applying to the browser maker first. You could view source, leave one service for another, or create something no platform owner had imagined. The app-store model put a gate back in the middle. Apple built the stricter gate. Google built the more permeable one. Either way, the gate came back.
Underneath all of this is a simpler feeling: I revere computers. I grew up programming them, repairing them, and building them from parts. My son does not yet have that relationship, and the machines we hand children now give them fewer ways to form it. I wanted his first pocket computer to leave some seams visible, some path toward understanding the machine in his hand. What I was considering instead was a magnificent computer sealed at the factory, in every sense.
I learned one thing in those years that I've never been able to unlearn: a wall is a decision someone made. It is not a law of physics. What gets built one way can be built another.
That is why one line from the Epic filings stayed with me. Apple's own people were on the record about why iMessage never came to Android. They could have built it. They chose not to. One senior engineer worried that doing so would "remove an obstacle to iPhone families giving their kids Android phones."
By then the obstacle had become my family. At the kitchen table, I could see exactly how it worked. Apple did not need to stop me from buying an Android phone. It only needed that choice to make my son's group chat worse. Once that was true, the parent would make the decision for them. The break between blue and green moved the decision from Apple to families, and the cost from Apple to kids.
The hard part is that I can see why Apple did it. If I ran the company, I would understand the instinct. You protect the thing that makes you different, and iMessage is one of the most effective moats ever built.
Scale changes the ethics. When you are the underdog, walling your garden can look like fair play. When your product becomes a generation's default social room, the garden becomes part of the public square. The move has not changed, but what it does has.
Apple does not have to enforce the wall directly. A mixed thread changes in ways everyone can see, and then the group does the enforcement. The harm has an author, and everyone knows who he is. The wall does not merely exclude one child. It recruits his friends to do it.
The cost does not appear in the sale. Apple can sell a phone. A family can solve a problem. The market records both as success. What it cannot record is the pressure that produced the choice, or the child who would have carried the cost of choosing differently.
Economics has a dry word for a cost pushed outside a transaction: an externality. In this case, the cost lands on the kid, the friend group, and every family split across two systems. The market has no price for it, which is exactly why it does not correct on its own.
For most of my life, I thought problems like this had exactly two homes. Either a company owns the thing and runs it, or the government does. Markets or the state. I argued about which one and never questioned the frame.
Then I found Elinor Ostrom, a political economist who won a Nobel for studying how communities govern what they share. Her work is new to me, and it has been rearranging my thinking ever since. She showed that shared things do not always need either a single private owner or the state. Fisheries, irrigation systems, forests: communities govern them through rules the people who depend on them write, test, and maintain together.
Her later work with Charlotte Hess turned to knowledge and information, which is the part that matters here. With messages, the danger is not depletion. You cannot overgraze a conversation. The danger is enclosure: someone walling off a thing that works better when it is shared. That is not a loose metaphor for what happened to messaging. It is the precise name for it.
Email is the clearest counterexample. Nobody owns email. No regulator runs it. Thousands of providers, from Gmail to a tiny host to a server in someone's closet, all speak the same open standards, and they all talk to one another. You can walk away from any of them tomorrow and keep every conversation. There is no green bubble in email, and no blue one either. There is no center to be outside of.
We already built the open version of messaging once. It works. It is still standing, decades later, carrying the world. The locked version is not the nature of the thing. It is a step back into a design we had already outgrown.
The wall has started to crack. Cross-platform messaging is better than it was. Photos survive more often. Reactions, typing indicators, and read receipts can cross. But the experience still depends on software versions, carriers, settings, and what everyone else in the thread has enabled. Even when the features cross, the bubble stays green. The machinery improves, but the mark of being outside remains. That is what grudging interoperability looks like: enough to reduce the damage, not enough to remove the boundary.
I bought him the iPhone.
I would like to tell you the decision was hard. The thinking was. The decision itself took about a second. You do not hand a thirteen-year-old a protocol war to fight in the hallways of middle school. You do not make your kid the price of your convictions. He needed to be back among his friends. The blue bubble was the way back. So I paid.
The phone worked exactly as promised. My son was relieved. My wife was relieved because he was. I was relieved too, mostly. By every measure the transaction could record, it was a success.
What I did not expect was what it would mean to him. The relief on his face was the kind you only see on someone who has been carrying something for a long time without saying so. He thanked me like I had handed him the keys to a car. And in a way I had. It was the threshold I crossed at sixteen, when a used car first let me reach my own life without a parent standing in the middle of it. He got there at thirteen, holding a slab of glass.
I still could not quite relax. The immediate problem had been solved, but the solution confirmed the larger one.
The size of his relief is the part I cannot put down. A boy should not have to feel freed to text his friends. Being let back into his own ordinary world should not arrive like a gift. All week I had precise names for the thing. Enclosure. Externality. None of them were any use at the kitchen table. The measure of his joy is the measure of the wall.
So who is supposed to build the open version? For a few days, I read my own decision as the answer to why it never gets built: because every one of us, in the moment, has a kid who just needs to get back in. That is true. But I have stopped hearing it as an indictment. A parent at a kitchen table is the wrong place to fight an architecture. The moment of decision is exactly where the system is strongest and we are weakest. Asking families to hold that line, one child at a time, is how the wall wins forever.
The line has to be held earlier, where the systems get made. That is builder's work, and it happens years before any particular kitchen table. I know because it is my work. What I am for is simple to say: systems that are interoperable by design. Protocols over platforms. Architectures where leaving remains possible, so staying has to be earned by making something genuinely worth using rather than by making your friends the switching cost.
The work I do now puts an open protocol underneath and a product on top. The protocol guarantees that you can walk out the door, which means the product has to be worth staying for. That is the whole bet. The commons is not a promise added after the product is built. It is part of the architecture.
There are two jobs on two timescales. As a father, my job was to get him inside, and I would do it again tomorrow. As a builder, my job is to make sure that trade is not sitting there waiting for the next father. Neither job excuses me from the other.
And I know what done looks like. Someday a thirteen-year-old texts his friends and feels nothing at all. No relief. No rush of freedom. No gratitude. Because there was never a wall there to notice.
Distributed via Edges.