Tech Brew Ride Home - Internet From Space with WSJ's Chris Mims

Episode Date: April 14, 2019

You might have noticed that I’ve gone down a bit of a wormhole lately when it comes to space tech, and actually, to understand why, listen to the very first minute of this conversation, because we actually work out why space tech has come to the forefront of my attention. Chris Mims had a piece in the Wall Street Journal this week about the suddenly race to deliver the Internet from space. This was triggered by the recent Amazon news of course, but also, listen for more deep dives into how and why space tech is suddenly having a moment. Chris’ article from the WSJ The CNBC article I refer to briefly Sponsors: Ethoslife.com techmeme.robinhood.com Learn more about your ad choices. Visit megaphone.fm/adchoices

On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco. Hey, who did this to you? What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. Welcome to a weekend bonus episode of the Tech Meme Ride Home. I'm Brian McCullough. You might have noticed that I've gone down a bit of a wormhole lately when it comes to space tech.

And actually, to understand why, listen to the very first minute of this conversation because we actually work out why space tech has come to the forefront of my attention. The we that I'm referring to is my guest, great friend of the show Christopher Mims. Chris has a piece in the Wall Street Journal this week about the sudden race to deliver. internet from space. This was triggered, of course, by the recent Amazon news, but also we get into more deep dives about how and why space tech is suddenly having a moment. My thanks to Chris for jumping on the show with me this weekend. I actually, I'm not sure why space tech has been coming up so much for me lately, even like, you know, the stupid black hole thing

this week. But in a way, it is starting to feel to me like it did a decade ago, and Tesla started making headlines for me, and then I started hearing about Google doing self-driving cars. So then all of a sudden I was like, oh, car tech is becoming a big thing. And now I sort of feel, for some weird reason, like space tech might be a next big thing. Are you sort of getting that sense a little bit, too? Absolutely. And obviously, this is, you know, an overnight success that's 50 years in the making.

There was a rustle of interest when, you know, Virgin started. pledging, you know, remember you could pay $100,000 and reserve a seat on a spaceflight, which I guess is sort of on schedule. You still can't take it, though. Right, right. Yeah, like, what I discovered, you know, I mean, I write about tech all the time, and I have not written about space until now. And I, it's exactly as it was for you. Like, I wasn't really writing about automotive. And then you kind of make this lateral step and you realize, oh my God, there's this whole universe, and we are at one of these unique times in history where a bunch of enabling technologies and a pile of money and certain demands are causing this

inflection point in this technology, and there's a lot going on all at once. Well, we're specifically going to talk about the idea of getting your internet from space, from satellites. And that definitely is not a new thing. I mean, going back 20, 25, 25 years at least, like there were things like Teledizic, Arridium, Global Star. I think Arridium and Global Star are still around. I know Aridium is. Vesat is still a going concern? Right. But so even though there are going concerns, it's still never really taken off. So do you have any sense of why all the sudden everyone's like, no, no, no, this thing that's failed kind of for 20 years. Now is the time for internet from satellites.

Right. So there are a number of unlocking technologies. First and foremost, Elon Musk has made it one of his many life missions to slash launch costs just per pound to a tenth of what they have historically been, and they've historically been quite stable. and then eventually to, you know, one hundredth of that. And, you know, he's doing that in ways that we all know about, like reusable rockets and smaller rockets and mass production of rockets

and, you know, new and innovative launch technologies and just stealing smart engineers and just not having the sort of legacy thinking of these defense contractors, which have gotten fat budgets from now, and from the Air Force forever to put up spy satellites and didn't have a lot of reason to innovate as a result. And we see where that led. The shuttle program was frankly a boondoggle. And until Musk gives us back that capability, we're sending up astronauts on, you know,

Soyuz capsules, which have the virtue of being extremely reliable. So part of reducing launch costs also has been reducing. the size of these rockets. And so then people are like, well, we're having smaller satellites. How do we make a virtue out of that? Because maybe they're less powerful or they can, you know, transmit fewer signals or whatever. And it turns out that what you can do with smaller satellites is put up a bunch more of

them and put them in a lower orbit, right? Because traditionally you were putting up satellites that were multi-ton, I mean, literally the size of a double-decker bus. I can't imagine the size of the rocket that is required to put a double-decker bus, you know, in geosynchronous orbit, which is like three Earth diameters away from the surface of the Earth. It's so far away. Right. Let me interrupt you because you actually really educated me on some pretty basic stuff. So geosynchronous satellites are 23,000 miles up, but low Earth orbit satellites are only 1,200 miles up. So the analogy that you use... Or less. Wow.

So in your piece, you're like, the analogy you use is if you picture a large naval orange, and that was planet Earth, then the height of a grape on top of that surface would be the equivalent of what a low Earth satellite would be. And actually, your piece has some great graphics on this too. So the link will be in the show notes, but check out the difference and what we're talking about here. And so obviously you can do smaller, cheaper. and then the advantage of the low Earth orbits, at least for the purposes of what we're talking about,

is latency because they're closer. Yeah, so the signal is you can bounce a signal in a tenth of the time or less, one web claims. And what that means is, you know, if you're bouncing a signal to geosynchronous orbit, that's a half a second there and back at the speed of light. And if you imagine a half-sumphant, Imagine if you were using your phone and you tap on something and a half a second later it reacted.

Or if we were talking and there's a half second delay, which there probably is anyway. Right, right, right, right, right. Like it would just be maddening. And then if you think about how complicated websites are, and especially if you think about encrypted websites, how there has to be a bunch of traffic back and forth, it makes it untenable. It's unusable. Now that said, there are startups like Estranus, which is an Andreson, Horowitz funded company that's getting a satellite up in 2020, and they're going to put a,

they're doing kind of a weird thing. It's a small satellite, but in a very distant geosynchronous orbit. And it will just give internet to Alaska. And Alaska is a place where if you want satellite internet, which for many people is their only option, you're going to pay like $300 a month for like two megabits per second up and down. So, you know, by cutting that cost significantly and uping the speed, they'll give you internet, even if it has long latency. Latency doesn't matter if you're streaming Netflix. Latency is okay for some amount of browsing, but you're never going to accept that level of latency unless you have to.

And so the key thing about geosynchronous orbit is, wait a minute, you're basically putting a cell tower, you know, up there, far enough, close enough that you can communicate with it directly. You know, what could that compete with? well, it could compete with your cell service. It could compete. If it's fast enough, it could compete with 5G and even broadband. Okay. Actually, I got a ton of questions about that in a second.

But maybe first, explain to me what my end experience would be. Is it like GPS in the sense that I would have a device with a chip in it that would just get the signal, bounce the signal back and forth? Or is it a thing where it's like the satellite bounces down to like a tower or something, like hotspotting, sort of like how cell networks do it? Well, this is an area where there's a lot of interesting innovation. And some of it I think we must remain skeptical of because lately, as you know, there's a lot of things that should, that people are promising and they just don't ever look

at the physics. This is why air power didn't work. The physics just don't work. This is why power at a distance doesn't work. So traditionally, if you want to get a signal from space, you got to put up a dish. still the cheapest and easiest way to do it. So, Astranes, for example, you just put up a dish, okay? And it's just aimed at that part of the sky.

It's like dish TV, whatever. But what is happening with these low Earth satellites is they pass overhead so quickly that you can either track them or you just aim a dish straight up and just the signal gets handed off as one satellite goes by after another. But they're traveling so fast. It's like over 12,000 miles per hour that every three minutes you're handing off to a new satellite. Another option is you can track the satellite as it goes across the sky.

Okay, now you've got to have a dish that's physically tracking that satellite, so it better be on your home, or you have what's called an electronically steered antenna. This is a newer technology. The ultimate goal is something the size of a pizza box, and it uses this really exotic technology. It's almost like an LCD screen, but it's not opaque to visible light. It is opaque to the wavelengths that these satellites are used. using and it actually kind of like filters or steers the signal somehow. So you can have an antenna that just sits out there like a solar panel,

but is actually electronically tracking the satellite as it crosses from one horizon to another, which again is really fast. Even if you're tracking it from horizon to horizon, it might be going overhead every seven minutes. Well, what about speeds? But again, all I want to know is can I dump Time Warner or whatever they call themselves now. like are we talking um because because don't currently when you get Wi-Fi on a plane it's using satellite too are we talking Wi-Fi on a plane speeds or are we talking like you know fiber speeds that's a great

question so it it's just like these 5G quote-unquote demos where it's like oh look how fast this 5G is and it's like right because you were a verge reporter standing under a tower and nobody else was using it at the moment like this is the trick of wireless is it's all about how many other people are trying to use it at the same time. So one web, which is the soft bank funded, largely soft bank funded, uh, satellite internet company, which is going to be the first to market with all this stuff with the new generation of this satellite internet. They've already launched six satellites. Their ultimate goal is 650. They said they might go to 1900, depending on demand. Their whole thing is, you know, on each of these medium-sized satellites, it has, I think,

16 beams, each of which carries 400 megabits per second. So you know, you do the math. It depends on how many people are in each of those beams, how much demand. It also depends on how many satellites you can see at that moment or how many satellites the ground station can see at that moment. Because you could get this satellite internet in two ways, at least. One is you could get it directly. The other way is they're just going to sell it as backhaul for remote.

cell towers, and you could just be connecting to the cell tower and never know that it's ultimately getting beamed to space, because it just happens to be easier to get that wherever you are in some remote part of the earth than to lay fiber to that tower. So, but then essentially what you're saying is, is if one of these, it proves to be super popular, then they would have the incentive and the resources to then, well, let's throw up more satellites and more satellites. So it is sort of like the idea of, well, you have some. slow connection at the football stadium because there's 90,000 people trying to use the same towers. But if it's successful, there would be a reason to throw up, you know, 50,000 satellites or something.

And keep in mind that this is apparently how Elon Musk plans to get to Mars. So he builds a satellite internet company that is large and profitable enough that it requires building zillions of his reusable rockets and launching them constantly and bringing the launch cost down and down and down, and throwing off enough cash to subsidize a mission to Mars. And Elon's is Starlink. Is that under SpaceX or is that a completely separate thing? I believe it is under SpaceX. They're pretty cagey about it.

The reason that we know what these networks will look like, other than companies like OneWeb that are public about it, is that they have to file with the FCC in the U.S. and the International Telecommunications Union, which is a UN body, because those are governing the spectrum that they're actually using to communicate with planet Earth. All right. So answer what was really my biggest question about this, which is, hold on. I've been told that 5G is right around the corner. 5G solves everything. We'll allow, you know, surgery across continents, self-driving cars,

and, you know, I'll be able to replace my time Warner, wired connection. So isn't this incredibly bad timing? If this is a huge capital-intensive investment to make, why would you make it right before theoretically 5G is going to come and make all this redundant? Right. So here I think a visual is appropriate. So let's go back to fruit, okay? Or let's talk about strawberries.

If you're talking about 5G, how do you get really fast signal to somebody? The radius of every one of those 5G towers, which isn't that big, sometimes they're like, oh, it's up to three miles. Like, okay, sure, if there's no trees in the way, or it's not raining or whatever, you know, the earth is an approximately round strawberry. Each of those 5G, you know, whatever they are, femto cells, the small towers that they need to enable this,

is like a seed on that strawberry. That's how much area it's covering. So if you want to cover the entire surface of the earth with 5G or the surface of that strawberry, it's got to be the most horrific mutant seed-studded strawberry. ever and you have to lay the physical backhaul for all that you got to install every single one of those little microcells you've got to get power to them you've got to get internet to them on fiber um so okay great if you're trying to you know add density to um you know time square if you are

if you're a verizon and you want to uh use 5g to connect the last uh you know half mile between your fiber hub suburb and all the homes in that suburb, rather than having to connect fiber to every single one of those homes. In those contexts, it makes sense in dense cities, et cetera. You know, adding it to existing towers, that's possible. But the reality is for most people, even just driving down the road is they're going to be constantly passing in and out of the effective radius of 5G radios and switching between 5G and what is 4G, or as AT&T likes to call it 5GE, which is just a marketing miss. Well, what satellite internet does for you is wherever you are, you are getting that speed.

And again, it's dependent on demand. But the whole attraction is you can very quickly blanket the earth because you're looking at the earth from outer space. So other players are, you know, Google's been talking about doing stuff like this, even with balloons at one point or whatever. And Facebook is always talking about doing stuff like this. And so to a certain degree, all of the big internet oligops are, basically, it's, I can imagine the spreadsheets where it's like, well, listen, if we increase the amount of people on the internet by 500 million, our profits go up by X percentage, you know. So like, it's literally just spreadsheet based economics here, which probably takes us to Amazon. So what do we think about Amazon and Project Kuiper?

I mean, you know, this could be an outgrowth of Jeff Bezos's midlife crisis. I don't know. I mean, they say it's early days so they don't want to talk about it. Obviously, you know, he has grand ambitions with Blue Origin and a lot of connections with other satellite companies because Blue Origin wants to be a launch platform for other satellite companies. no direct connection to Blue Origin yet. But, you know, I feel like where we're at with all these companies, you know, especially Facebook and Amazon exploring this, is where we were a few years ago when everybody was thinking about building their own

self-driving car, including Apple, and then later they abandoned it when they realized how hard it was or that it just wasn't for them. Like, I think a lot of companies are going to make a swing at this. But what, you know, what analysts who look at the financial side say is like, look, they can't all launch this many satellites. There isn't enough money or demand. So I think what's going to happen is that's one reason that SoftBank has made such a big bet, is they're like, look, it's like the ride hailing market.

Whoever gets in there first sucks up the demand, sucks up the capital, and ultimately wins whether or not their technology is the best. So I think a lot of them are going to try. I mean, I don't know where it's going to go. Google has already pivoted where in 2014 there were these articles about Google working on their own satellites. and they apparently haven't done much on that and instead are taking like the technology from Project Loon, which is their balloon-based internet and licensing it to a Canadian satellite

company to establish lateral connections, you know, a mesh of connections between their satellites up in space. So, you know, not all these companies are going to stick with this. Some of it is just pure hubris. I have no idea how determined. Amazon really is in the end. Well, I've been reading around on this, and I actually, all of a sudden, I'm quite bullish on this is, this is maybe potentially another AWS for Amazon. So hear me out for a second. There's a great, hopefully I'll remember to link to this in the show notes. There's a, there was a great CNBC piece this week talking to like a dozen people in the space industry

who are kind of thinking of this as earth-shattering. You know, everyone's like, well, Amazon's getting into healthcare, and so all of a sudden you're seeing all the insurance companies merging and things like that. But, the space industry people were like, Amazon is a dead serious player. And they were pointing out all of the different synergies. Like, first of all, it's super capital intensive. Like, I think Aradium's constellation costs $3 billion to put up. So someone was like estimating that it might cost, it might cost between $3 to $5 billion for Amazon to put up its own constellation. Well, if anyone has a ton of capital to spend, it would be,

Amazon. If anyone is willing to lose a lot of money investing for years to pay off down the road, it's Amazon. And then they have the infrastructure because then Bezos owns Blue Origin. So all of a sudden, great, Blue Origins got this deep-pocketed customer. Amazon quietly announced something called the AWS Ground Station, which is some sort of satellite facilities around the world. I talked to the head of that. He's the guy who gets the last word in my article. And that was fascinating. because he's a real veteran. So he has as a veteran of, you know, defense contract era satellite companies,

but he really knows his stuff. He knows outer space. He knows what it takes to build a satellite. He knows what it takes to get one into space and how difficult it is. But he is, even knowing all that, you know, he started on his own initiative. He got hired as an engineer at AWS,

and he's like, ah, I'd like to use my satellite background. can we build a bunch of ground stations and connect them to AWS? And they were like, sure, do that on your own. And now he has a team of like 30. And basically what he's saying is on ground stations, which communicate with the satellites in space or this critical piece of infrastructure that you don't want to have to own. You share it.

That's why it's part of AWS. It's a fundamental principle. And his argument is, you know, as launch costs fall, he thinks that you're going to get people, as he put it, building satellites in a garage. and then they pay somebody like Virgin Orbit, which launches satellites on these little rockets from under the wings of a plane

to get them into space, and then they connect them to AWS Ground Station. And all they had to do was build that outer space-proof satellite. That's still hard, but they're commoditizing space is the idea. And so I agree. I think there are a ton of synergies here. I think another synergy that's not as often appreciated

is that a lot of these satellites, that get put up are useful for logistics. So you want to track a ship or you want to track a shipping container or some or a truck or whatever. You know, Amazon is making huge investments in competing directly with UPS and FedEx with its own planes and stuff. Like, well, those those drones need to be directed. Those drones that are going to fly over and deliver the boxes need to be direct. It's true. It's true. And and so, you know, people have done a lot of handwavy things about like, well, the internet of things is going to need a lot more bandwidth. It's true, though. but that's a chicken and egg problem, right?

It's like, where do you get that bandwidth? Is it consistent everywhere? Like, getting it from space kind of cuts the gaudy and not. It gets you a lot all at once, and it gets you the potential for a lot of applications that I think we can only guess at now. Well, and then Amazon would benefit, AWS would benefit from that. All right, let me give you two more things as I've been poking around on this. The idea that I keep reading articles that, like, you know, maybe test. is going to have problems raising more capital. SpaceX also might have problems raising more capital.

So what if SpaceX stumbles? They're in Seattle, Amazon's backyard. If there's a stumble, all of a sudden there would be a ton of smart engineers that Amazon could quickly hire and no one has to move or change schools or anything. And then the other thing is, so owning the home seems to be, like, I keep imagining this future where we're going to pay a monthly Amazon tax, they'll call it, you know, membership and prime or whatever. But so why not your internet service be a part of that? Why not that be a part of your big prime subscription? And then, by the way, delivering internet to the home, well, that helps with things like

Prime Video to deliver the $250 million investment in Lord of the Rings and stuff like that. Like there's just, there's a thousand different ways that I think that there's synergies here for Amazon. Yeah, I think that's true. I mean, don't forget that, you know, everybody wants a piece of your home internet pie, because, you know, it has been, it is a local monopoly, right? And so that is why, you know, Verizon wants to use 5G to compete with Comcast, Semptime Warner.

So, you know, Amazon, if they can leapfrog all of that nonsense and do it from space, like absolutely, I could see them being your ISP as well. I mean, I think, I mean, look, it'll either all flame out like the fire phone or 20 years from now, you and I will look back and be like, wow, what a prescient conversation we had when we were just kind of wildly speculating about where this could go.