The Cell Tower Is the New Landline

Welcome. I think it's time to question everything.

And I think there's a decent chance that someday you're going to look back at your cell phone bill the same way you look at a landline bill today. If you even have one, because I'm starting to wonder if the cellular network, at least the way we currently think about it, is already becoming a bridge. Now, before somebody who works for a price and starts typing in an angry comment, I'm not saying that we're tearing down every cell phone tower next Tuesday.

I'm not saying satellites can handle every phone on Earth today. They can't. I'm saying something much more specific. I don't think people actually want cellular service. We want connection. And those are not the same thing. Take this thing here. We still call it a cell phone. Think about that name. Cell phone. How much of what you actually do with this thing has anything to do with making a traditional phone call?

I watch video, I send messages, I use, I navigate, I upload things, I listen to music, I send pictures, do work, FaceTime, browse the internet. I do scroll like everybody else does. Pretty much everything that makes this device useful to me is data. Hell, even when I make a phone call half the time it can travel over Wi-Fi.

So what exactly is this thing? Because I don't think it's really a cell phone anymore. It's a computer. It's a little computer that I expect to be connected everywhere I go. And once you look at it that way, I think the question gets a lot more interesting, because if what I actually need is data, why do I care how the data gets there?

That's the human need connection. Not a tower, not 5G, not LTE, not Wi-Fi, not satellite, not whatever the hell. Six G eventually means. I just want this thing connected, which means the network itself. Maybe another one of these interfaces we confuse with the destination. For decades, the cellular network was an incredible solution to a really difficult problem. How do you keep somebody connected while they move around?

You can't drag a wire behind them. So we divided geography into cells, put antennas on towers, connected those towers back to a network. And as you moved around, your phone moved from one cell to another. Tower to tower. Handoff to handoff. And when it works, you don't even notice. Which is pretty amazing when you think about it. There's a gigantic invisible machine working around you all day.

Just see, your stupid little TikTok video doesn't stop when you drive through town, but then you leave town for hours, then three, two, one, and then nothing. And suddenly you become a telecommunications engineer. You're holding the phone near the window. Maybe if I put it up here. How about over here? What if I turn it sideways? Like somehow your body has been blocking the Verizon network for the last 20 miles.

We've all done it. And for decades, a dead zone made complete sense. Because if the tower couldn't reach you, the network couldn't reach you. Pretty simple. But that limitation wasn't really technological, at least not entirely. A lot of it was economic. Think about what it takes to cover one piece of geography. Somebody needs the land or lease permits, tower backhaul, antennas, equipment maintenance, upgrades, people who can go fix the thing when lightning blows the shit out of it.

All of that costs money. And if that tower serves 50,000 people, then fine. But if it serves 50, that mass starts getting ugly. If it serves nobody except three hunters and a raccoon. Nine months out of the year. That's a problem. Good luck getting that tower built. And that's why the maps have holes. Not because humans don't know how to build another tower.

Because at some point, building another tower isn't worth it. The network has always been constrained by geography. Except now the tower can move and it can be hundreds of miles above your head. That is the part where I think people need to pay attention to direction again, because direct satellite connection to regular phones isn't science fiction anymore. It's happening today.

Phones people already own can communicate with satellites. No giant antenna, no special phone with a six foot stick hanging out of it. No briefcase, no satellite dish strapped to your head. Your phone. And yes, today there are limitations. Some satellite phone features are mostly messaging, some support, limited applications. You generally need need a reasonable view of the sky.

Buildings are certainly a problem. Heavy tree cover can also be a problem and the capacity is nowhere near close to what we get from a dense terrestrial network. And if you're sitting in the middle of Manhattan, I'm not suggesting we replace every antenna in New York City with one satellite flying over new Jersey, because that's just stupid. But I keep coming back to the same thing I said in the first two memos.

I'm not particularly interested on whether this replaces the cellular network. Today I'm interested in the direction because right now the direction is more satellites, more phones, more coverage, more applications, more data, better antennas, better irrigation integration, more capacity, and fewer places where network simply disappears. So I don't think the right direction is is satellite as good as 5G today?

Of course it isn't. The question is, what happens if it becomes good enough?

Remember Kodak early digital cameras were shit compared to film. They didn't save film. Remember Kodak? Early digital cameras were shit compared to film and that didn't save film. Remember, streaming early internet video was terrible. That didn't save blockbuster. Remember software? I screwed things up today. That doesn't tell us where the bridge is headed. Technology does this over and over and over again.

Bad and useful and cheaper. Easier. Good enough. And then suddenly, the thing that everybody told you could never replace the old thing becomes normal. And satellite has one unbelievably important advantage. It doesn't care whether you're economically interesting enough to deserve a tower. The satellite is already moving over the mountain from the highway, the desert, the farm, the ocean and the town with 30 people in it and the place nobody lives at all.

That changes the economics of coverage. Maybe not capacity, but coverage. And I think coverage itself could start becoming abundant. That's a really big deal because for basically the entire mobile era, coverage has been something companies sell us. Look at our map. We cover more people, we cover more highway highways. We work here. They don't work there. This carrier works at my house.

That carrier works at your camp. We've built entire purchasing decisions around which company happen to put enough towers near the places we spend our lives. But what happens when those holes start disappearing from the map? What happens when the question do I have service here starts becoming strange. And this is where I think the landline analogy matters. Landlines didn't vanish because they suddenly stopped working.

They worked great. They still work in technology wasn't necessarily bad. Something else happened. The landline lost its monopoly on the human need. At one point, if you wanted a telephone in your house, you needed that physical connection. That wire was the service. Then mobile arrived and suddenly telephone service didn't require the telephone line anymore. The old infrastructure could still exist.

It just stopped being the assumption. And I wonder if the cell tower is about to have its own landline moment. Not because towers disappear, but because the tower loses its monopoly. Mobile connection. And maybe I'm not even being aggressive enough because there's another bridge inside this. The cellular plan itself. Think about it. Why do I buy cellular service?

I want my device connected. Okay. But imagine this device 5 or 10 years from now, I'm sitting inside, outside my house. It's connected to my local network. I walk outside. Maybe terrestrial cellular is fastest. I drive 20 miles into the middle of nowhere. Satellite. I walk into a hospital, a factory, an airport, an office building. Maybe it uses private.

A private terrestrial network inside the building. Maybe Wi-Fi, maybe something else. Why am I choosing any of that? Why do I even need to know? Let the device choose. Better yet, let intelligence choose what connection is available. Which is fastest, which is cheapest, which uses the least amount of battery? Which is secure? Which? What latency does this application need?

What capacity is available? Use that one and if conditions change. Switch it.

Invisible. This is the same thing we just talked about. Software. I don't want to operate the application. I want the outcome. And maybe I don't want to operate the network either. I want connection. So just like software may become invisible, the network may become invisible, you know, by Wi-Fi, satellite, cellular, whatever. Maybe eventually you just buy connection. Or maybe connection gets bundled so deeply and everything around us, you barely think about buying it at all.

And before somebody clips that sentence and says, this idiot says Verizon is going bankrupt. No, I don't know that. Actually, the companies that understand this transition may do extremely well because this is already happening through partnerships. Terrestrial carriers and satellite companies are working together, which is interesting by itself, because this tells me that people who currently own the bridge can see the next bridge.

It doesn't mean they lose. Kodak could have won digital photography series could have one. Internet commerce. The problem wasn't that the next bridge appeared. The problem was failing to understand what business you were actually in. So if you're a cellular carrier, are you in the tower business? The spectrum business, the cellular plan business? Where are you in the connection business?

Those lead to very different decisions. And towers still have enormous advantages. If I have 80,000 people in a football stadium all trying to upload a video, I'm at capacity close to them. If I'm inside a concrete hospital, I probably need infrastructure inside the building. Factories, cities, apartment towers, airports, subways, dense neighborhoods, high speed autonomous systems. There are a million situations where terrestrial infrastructure makes complete sense.

The physics didn't go away. So maybe towers become the dense layer, the high performance layer, the indoor layer, and satellite becomes the blanket. The thing underneath the whole system where the tower stops. The connection doesn't.

I'll think about what that does beyond the phone, because this is where I think this gets much bigger than whether I can send a text while I'm camping. Phones are just the obvious example. What else needs connectivity? Cars. Trucks. Ships. Planes. Drones. Farm equipment. Sensors. Pipelines. Construction equipment. Emergency systems. Machines. Robots. AI. Every intelligent machine we build eventually runs into the same problem.

It can be incredibly smart, but if it can't communicate, if it can't receive information, if it can't send information, if it can't update, if it can't coordinate, then its usefulness has a geographic boundary. Connectivity removes part of that boundary. Imagine an autonomous machine that can operate anywhere. It has a connection a vehicle, a robot, a drone, some machines we haven't invented yet.

Putting a city connected. Put it on a farm. Connect it. Put a desert connected. Put it halfway up a fucking mountain and it's connected. Now suddenly that machine isn't just doing work. It's producing information. Real world information. What did it see? What happened? What failed? What worked? How did that road behave? What did the weather do? What did the machine encounter that it had never encountered before?

The information can go back to intelligence. And the intelligence gets better. Then the better intelligence goes back to the machine. See what's happening. Software needs compute. Computing needs energy. Intelligence needs data. Machines create data. Machines need connectivity. Connectivity needs infrastructure. Satellites need to get into orbit. Orbit requires launch. Launch requires me. Factory manufacturing requires energy. Everything starts connecting to everything else.

And this is the part that I don't want to jump over too quickly. Because in the last memo, I said that if digital things become abundant, like software, images, video, music, maybe even intelligence, then economic value probably starts migrating toward the things that remain scarce energy, compute, Silicon networks, factories, machines, physical infrastructure. Now add something else to the list the ability to connect it all together.

And once I started looking at connectivity this way, something else started bothering me. Because Starlink isn't just a satellite brand. Actually, Starlink by itself is already kind of a weird brand to think about. How do you compete with a giant satellite network? Well, first you need satellites. A shitload of them. Then you have to manufacture them. Then you have to put them in space repeatedly and affordably, which means suddenly your internet company has a rocket problem.

Starlink works differently because it's embedded inside of space. The network and the launch platform are connected. That matters. Now keep going. Once you have the network, what benefits form global connectivity? Cars. Autonomous systems. Robots. AI machines operating in the physical world. Real world data. And suddenly I find myself looking at the same names again. SpaceX Starlink tech Tesla x I optimize energy, launch satellites, artificial intelligence vehicles, robots.

And I keep wanting to put them into separate boxes. Rocket company, internet company, car company, AI company, robot battery, social network, whatever. That's how we normally understand companies. We name the product and assume that's the business. Sound familiar? Kodak sold film. Except film wasn't a business. Sears had the catalog. Except the catalog wasn't. The business software has applications except maybe the application isn't the business.

And now I'm looking at this collection of companies and wondering if we're really doing the exact same thing again. Looking at each individual product, instead of asking what the whole system is trying to accomplish. Now, I need to be really careful here because the internet is completely fucking insane when it comes to Elon Musk. There are people who think the man is going to single handedly save humanity, and then there are people who think he wakes up every morning trying to personally destroy it.

I'm not interested in either religion. This isn't Elon Musk is a genius and it isn't. Elon Musk is an idiot. I don't give a shit. That's not the interesting question. The interesting question is why do all these pieces keep fitting together? Because maybe they don't. Maybe I'm forcing a pattern into a bunch of independent businesses. That's possible. Maybe Tesla is just a car company, and Starlink is just an internet service.

Maybe I is just another AI company. Maybe Optimus is just a robot project. Maybe SpaceX is just really good at rockets. And all of this happens to sit around the same person. Could be. But if this series is about questioning bridges, I think we at least have to look. Because what if the individual companies are the wrong unit of analysis?

What if the interesting thing is the ecosystem between them? A company that builds machines. A company that builds intelligence. A network that connects the machines rockets to deploy the network, energy systems that power pieces of it. Factories that manufacture real world machines that generate data which can improve the intelligence, which improves the machines. This starts looking less like a collection of companies and more like a system.

And if it is a system, we need to understand what the hell is building. Not because Musk says so, not because his fans say so, not because his critics say so, but because the pieces are sitting right in front of us. And I want to know whether they actually connect. So in the next memo, we're going to stop looking at companies one at a time.

Tesla, SpaceX's Starlink x AI, robots, energy, maybe even AKS. And instead we're going to put the whole thing on the table, because I think there's a possibility that the company everybody thinks they understand isn't actually the company. Maybe the real company doesn't have a name yet.

I don't know where this bridge ends, but I think it's worth finding out. Goodbye.