Why Elon Musk is really building Starlink

T-minus three, two, one, lift off. That rocket is part of a billionaire's plan to put your internet in space. Jeff Bezos' Project Kuiper. No, no, the other one. Elon Musk's new Starlink satellite. One web. Backspy. Richard Branson. Keep it all straight. Yeah. Hold on. Why are multiple billionaires trying to put the internet in space? And where is the internet now? A new space race to provide internet access at fast speeds to the world. You're Elon Musk. I am? Well, finally, it is here. Starlink. Jeff who? From Musk to Bezos to Branson. Let's get ready to rumble. All right, I'm going to explain this great internet space race. This is especially relevant right now because as we were making this episode, Elon Musk announced that Starlink is available in Ukraine. More on that in a minute. These systems could change how millions of people get access to the internet. They could be huge, if true. You want to see something cool? This is a piece of the internet. It's an unused chunk of the over 800,000 miles of cables that crisscross our oceans. Here, let me show you. There they are. As of 2021, there were roughly 436 of these cables in service. They carry the vast majority of international internet traffic today. Some estimates put it up to 99%. Meaning, for most people, right now, your internet isn't in space. It's underwater. The goal is to get information from one computer to another. And you've probably heard that all of the information inside of your computer is binary, right? Ones and zeros, on or off. These cables carry that information as flashes of light. Imagine... Hold on one sec. Imagine turning a flashlight on and off in code down a tube. I'm definitely not going to shine it right at the camera. Now imagine doing that millions of times faster down a tiny glass fiber across an entire ocean, and you've got the right idea. That's pretty cool. Yeah, humans are wild. I wanted to know how on earth we made this internet infrastructure, so I toured a ship that lays these cables as part of a Vox show. That's the internet! Look how excited I am. And here I am finding out the answer to my question. Ships like this load up cable like this and then slowly cross the ocean. We're talking like six knots, which is seven miles an hour. This takes weeks. Dropping the cable off the back of the boat. Weaving a literal web around the world. Which means that firehose-looking things like this really are just laying there on the bottom of the ocean. And you're using them all the time. This breaks my brain. Like, this YouTube video that you're watching right now is one tiny drop in the unthinkably huge stream of things that people are saying to each other all of the time, all across these cables, and they're all made of light. Like, if you were to describe to me an alien species, and you were like, they communicate incredibly complicated audiovisual information to each other using only light and these light boxes that they made for each other, I would be like... My biggest question when I heard this for the first time was, aren't there parts of the ocean that are deeper than Mount Everest is tall? If you're using cables as opposed to satellites, how do they get across those parts? Just like, look at this. This is the deepest ocean trench. Some of these cables have to go over some of these super deep parts. I actually couldn't find the answer to this question until I reached out to one of these ship-laying companies and asked if this was a big problem. Like, would a cable hang through the water? Or do you build bridges? And they actually got back to me to say, no, we don't build bridges. And no, we don't let a cable just hang through the water. They said they really do lay enough cable so it does fall all the way to the bottom, which can get difficult and expensive, which is why they try to avoid the deepest parts. And so far, they say that they have. I've made videos about this before. Like, a lot of them. And every single time, people are shocked. They say they thought the internet was in space. And that makes sense. Think about the language that we use. The cloud upload, download, download all his nudes. One time I asked people to draw the internet, though some of you awesome nerds drew very detailed infrastructure drawings. Most people thought of the internet like this. Except this person. I thought of the internet as a bunch of cats. That's my favorite. Anyway, my point is that our existing internet infrastructure is a marvel of human ingenuity. So why do Elon Musk and Daddy Bezos want to put it in space? History is about to be made in the science of communication among men. The idea isn't new. This is Telstar 1, the first communications satellite, launched in 1962. This understanding which will inevitably come from the speedier communications is bound to increase the well-being and security of all people. Mixed results. With fiber cables, we can turn 1s and 0s into visible light. With satellites, at some point, the information has to go up and down. So we turn those 1s and 0s into radio waves. The question that I always had is, how do 1s and 0s become waves? The way to think about this is, you could change the amplitude or the height of the wave so that every tall wave is a 1 and every short wave is a 0. So like 1, 1, 0, 0, 1, 1. Or you could change the frequency of the wave so that you could say every long wave is a 1 and every short wave is a 0. 1, 0, 0, 1. It is way more complicated than this now. Companies combine methods, they develop new ones. That's the right idea. No matter what, though, to communicate via satellite, radio waves with binary information have to make this trek. It's a big limitation for satellite internet. So far, this, cables carrying visible light, has generally been faster, cheaper, and more reliable than this, satellites using radio waves. The file's not loading. Nothing's loading. It's still loading. To understand what we're comparing, it helps me to think about a water pipe. Bandwidth is the width of the pipe, how much can go through in a set period of time. Like this, megabits per second. Just to really tie this together, a bit is one of those 1s and 0s. So 400 megabits is 419,430,400 1s and 0s per second. Damn. Latency is like the length of the pipe, the time it takes the information stream to travel from one end to the other. It's measured in milliseconds. Low bandwidth, high latency, worse internet. High bandwidth, low latency, better internet. Satellites before these recent projects, fiber cables. And satellites would cost you more. But, and this is a big but, that might not be true forever, and it's not true everywhere now. The biggest reason that people are so excited about all of these new satellite internet projects is that though our existing internet infrastructure is awesome, it's not enough. Though, if you're watching this, you might have really good internet. Lots of people don't. In some places, it's prohibitively difficult to lay cables. In others, crucial ground-based service might be threatened by politics or war, which is happening in Ukraine right now. I'll come back to this. And during peacetime, companies might decide that it's not financially worth it to lay cable because there aren't enough people who would pay for their service. Think about roughly the same amount of cable in a city with hundreds of people paying for it versus in a rural community with much fewer. And if they do lay cable, they might charge even more. And that right there is a huge problem. This map shows the share of each country's population using the internet. Look who's able to get online and who isn't. I'll put a link to this in the description if you want to explore it. This difference is called the digital divide. And it's not just between countries, but also within them. It's really hard to get internet here. We do not get internet service. There was no high-speed internet. Take a look at this. This is a map from the U.S. Federal Communications Commission, or FCC, showing where they say about 14 million people don't have access to fast internet. They're in those light parts. But look how much lighter this map is. It's from Microsoft, and it shows a very different estimate. Over 120 million people. The one by the FCC is based on data that is self-reported by internet service providers. And it gets worse. The form that the FCC gives them, Form 477, asks if, quote, they are providing or could provide broadband service in an area. So, for example, these are New York City's census blocks. If an internet provider is giving anyone in this block high-speed internet, they get to count the entire block. There are nearly 800 people living in this block. So, the digital divide is a big problem globally, between cities and rural communities, and within communities themselves. And the data probably wildly underestimates the problem. It's a problem that we need to solve, given how important the internet is for just participating in society today. It's really two problems wrapped into one. It's accessibility and affordability. This is why the great internet space race actually matters. There's a vision here of a world where your ability to freely participate depends a lot less on where you live and how much money you can shell out for internet. And that's what these companies say they're trying to solve. Estonic will serve the hardest to serve customers. The goal here is broadband everywhere. To the roughly 4 billion people on earth who don't yet have it. Access to broadband is going to be very close to being a fundamental human need. The cynical side of me thinks these companies don't operate out of the goodness of their own hearts. So why do they want to be the one to bring internet access to the world? And the answer is that bringing internet access to more people is a great thing to do. And it can be a great business. Providing broadband is more like probably 30 billion a year. If a company can use satellites to make the internet better, cheaper, and more widely available, they could get as customers the millions of people that are underserved by existing internet service providers now. And this is where it gets really cool. They're going about the problem differently than satellite internet companies that have come before. Older satellite systems like HughesNet and Viasat put their satellites really high up and keep them stationary over one spot on the surface, which lets them cover huge areas. The HughesNet Jupiter 2 satellite covers all of the continental United States. But when you're up that high, it takes a really long time to send those radio waves that we talked about back and forth, which means that your latency is very high. If you brought those satellites lower, you could reduce that time. But you'd need more of them to cover the same area. And to stay at that altitude, they'd need to orbit. This is what all of the big projects that you've heard about are doing, though they're at different stages. Project Kuiper hasn't launched yet. OneWeb has, but was rescued from bankruptcy in 2020 and is focusing primarily on businesses, not individuals. In this race, Elon Musk is winning. Starlink is here. We have the magic space Wi-Fi. That's right. Internet from the stars, baby. Let's take a look at what's inside. Here we go. We've got a satellite dish, a router, and fast internet. They're just two instructions, and they can be done in either order. Point at the sky, plug in. Voila. Not gonna lie, Starlink is already shockingly high bandwidth and low latency for a satellite system. It seems to be somewhere in the middle on cost, but these comparisons aren't great because fiber isn't everywhere, which is the whole point. And Starlink is still in its infancy, which you can kind of see in other ways, too. Everything was going fine until I encountered its ultimate nemesis. Trees. That red is where it's being interrupted. That sucks. It could get a lot better, though. Starlink uses visible light between satellites, like the fiber cables in the ocean, which means that these satellites actually have an advantage over cables. You've probably heard that the speed of light is constant, but that refers to light in a vacuum, like space. It's slower in other materials, like a glass cable. The optimistic take here is that Starlink could challenge not just other satellites, but fiber itself, the goat. The more likely scenario is that Starlink wins in some places, but fiber still wins in the vast majority. But just look at this. This is a site that tracks where Starlink satellites are already. This is so cool. But if you're an astronomer, for example, looking up at the stars, these satellites can really mess up your science. I think it's really a threat to send not a couple of hundred satellites, but having 10,000, 20,000, 50,000 satellites in no Earth orbit. This is a really big deal. And even if you like Starlink, it shouldn't be dismissed. Ground-based astronomy is crucial to answering our questions about the universe. And if that's not enough for you, then it also includes planetary defense against asteroids. My take is we need to give the remaining 4 billion people on the planet Internet access. And at the same time, we can't let looking at each other prevent us from looking up. We need to find ways to do both, whether that's more data on exactly where these satellites are so they can be avoided, or more rules about where they can be. These companies have a real shot at giving people around the world They're already coming to the rescue in moments of crisis. Elon Musk just announced that Starlink is active in Ukraine. And if Russia doesn't like it, they can shake their fist at the sky. There's something extremely powerful happening here. In general, if what these companies provide ends up better and faster, so higher bandwidth and lower latency, but not cheaper, well, it's mostly left for those who can pay for it. High-frequency traders, wealthy people, fine, but not world-changing. The real impact happens when we use technology to make things better, faster, and cheaper, to turn luxuries into commodities, to take things that wealthy people had access to and make them available to everyone. We've done this before with everything from refrigerators to iPhones, to the Internet to some extent. So can it continue? And what's the impact if it does? Research shows that where the Internet goes, job growth follows. More new businesses get formed. Fewer people are unemployed. People have better health outcomes, better access to education. We saw all of this to an extreme when the pandemic started. New satellite systems could be a huge part of this, along with more investment in the systems that most of us rely on right now. Our Internet infrastructure is a monument to human creativity and hard work and the desire to connect. It blows me away, but it's not finished.