WEBVTT - Samanth Subramanian on the Undersea Cables That Keep the Internet Alive 0:00:02.720 --> 0:00:07.200 Bloomberg Audio Studios, Podcasts, radio News. 0:00:18.079 --> 0:00:20.920 Hello and welcome to another episode of the Authoughts podcast. 0:00:20.960 --> 0:00:24.360 I'm Tracy Alloway and I'm Jill wi isn't Joe. I 0:00:24.400 --> 0:00:26.640 am going to send you a link. Okay, you're going 0:00:26.680 --> 0:00:29.360 to open it on your laptop, which we have expressly 0:00:29.400 --> 0:00:31.480 for this purpose, and tell me what you see. 0:00:31.640 --> 0:00:33.879 Okay, Okay, I'm pueing it. This is a fun experiment. 0:00:33.880 --> 0:00:35.280 We've never done anything like this before. 0:00:35.320 --> 0:00:37.760 I know, we've never been so prepared for a podcast 0:00:38.240 --> 0:00:39.640 as to have a link ready to go. 0:00:39.840 --> 0:00:43.440 It's a map of undersea case. It's actually the Encyclopedia 0:00:43.440 --> 0:00:47.320 Britannica page of undersea cables, and there's a really old 0:00:47.360 --> 0:00:50.480 fashioned looking map here. Looks pretty cool. 0:00:50.840 --> 0:00:52.000 What does that remind you of? 0:00:52.960 --> 0:00:53.840 Oh no, I lost them. 0:00:55.560 --> 0:00:55.960 I knew this. 0:00:55.920 --> 0:00:58.880 Wouldn't wor shipping lanes. Yes, okay, yes, that's I got 0:00:58.880 --> 0:01:00.000 the right answer. Yeah. 0:01:00.120 --> 0:01:01.720 So I've been looking at these maps and they look 0:01:01.800 --> 0:01:06.240 remarkably like shipping lanes, even though upon further reading, it 0:01:06.280 --> 0:01:10.120 turns out that subc cables do not precisely follow actual 0:01:10.240 --> 0:01:12.720 like major shipping lanes. But since most of them are 0:01:12.760 --> 0:01:16.480 going to and from a major economic center a city 0:01:16.600 --> 0:01:19.160 or something. They look a lot like shipping lanes. And 0:01:19.200 --> 0:01:21.920 the reason I bring this up is because I think 0:01:22.520 --> 0:01:26.319 with everything technological nowadays, you know, whether it's AI or 0:01:26.600 --> 0:01:29.200 the basic Internet, there's a tendency to think of it 0:01:29.240 --> 0:01:32.959 as this very ephemeral digital presence in our lives. But 0:01:33.080 --> 0:01:36.400 of course, as we've been discussing on a couple podcasts recently, 0:01:36.480 --> 0:01:40.240 there is an incredible physical architecture which is the source 0:01:40.280 --> 0:01:43.560 of all these things, whether it's subc fiber optic cables 0:01:43.560 --> 0:01:46.399 that kind of look like shipping lines, or massive data 0:01:46.440 --> 0:01:49.720 centers that cost a lot of electricity and commodities to produce. 0:01:50.520 --> 0:01:53.560 Undersea cables is one of the most fascinating things to 0:01:53.640 --> 0:01:57.000 me because there's been a cable right between London and 0:01:57.080 --> 0:02:00.320 New York for over a hundred years. I know that 0:02:00.320 --> 0:02:04.040 blows That's actually why the UK pound sterling rate is 0:02:04.040 --> 0:02:07.200 called cable, That's why it's nicknamed that. It blows my 0:02:07.400 --> 0:02:09.760 mind that, like over one hundred years ago, whenever it 0:02:09.840 --> 0:02:11.679 was when they first had the telegraph, that they were 0:02:11.720 --> 0:02:15.040 able to string a continuous line from London to New 0:02:15.080 --> 0:02:17.120 York like that blows my mind that they could do that. 0:02:17.320 --> 0:02:20.720 Whatever that was like a century ago whenever, But it 0:02:20.760 --> 0:02:22.519 blows my mind that they can still do it today, 0:02:22.919 --> 0:02:26.000 like it actually it hurt. I still find it hard 0:02:26.040 --> 0:02:29.200 to believe that they can actually have one length of 0:02:29.320 --> 0:02:32.720 cord that goes under sea this far, even with all 0:02:32.720 --> 0:02:34.840 the technologies, I still can't wrap my head around how 0:02:34.880 --> 0:02:35.280 they do that. 0:02:35.440 --> 0:02:37.480 Yeah, and some of those cable lines I think are 0:02:37.600 --> 0:02:40.080 armored in various ways to protect them. They're talking about 0:02:40.120 --> 0:02:43.120 doing like more land fiber optic cables now as well. 0:02:43.520 --> 0:02:46.400 There's a lot that we should discuss, not just because 0:02:46.440 --> 0:02:49.080 this topic is incredibly interesting and we've been meaning to 0:02:49.240 --> 0:02:51.480 for a long time, but also because with all the 0:02:51.919 --> 0:02:56.320 geopolitical volatility that we're seeing nowadays, you always hear subsea 0:02:56.400 --> 0:02:59.800 cables coming up as a potential source of vulnerability totally. 0:02:59.840 --> 0:03:02.000 You hear about attacks and them from time to time, 0:03:02.040 --> 0:03:04.280 and I guess there's probably some redundancies and other ways 0:03:04.320 --> 0:03:07.040 to route around, but it is weird to think that, 0:03:07.080 --> 0:03:10.560 like in theory, you could just like cut off some 0:03:10.800 --> 0:03:12.840 part of the world from the Internet, like clipping all 0:03:12.880 --> 0:03:14.079 the cables at least I get them. 0:03:14.240 --> 0:03:17.320 You could send out little drone lobsters and they like 0:03:17.440 --> 0:03:20.480 cut the cable in the deep sea bed with their little. 0:03:20.240 --> 0:03:21.880 Lobster clubs are literally happening. 0:03:21.960 --> 0:03:24.280 Okay, we should actually speak to someone who knows about this. 0:03:24.520 --> 0:03:26.880 We do, in fact have the perfect guest again, someone 0:03:26.880 --> 0:03:28.760 we wanted to speak to for a very long time. 0:03:28.840 --> 0:03:31.200 We're going to be speaking with some month, Supermanian. He's 0:03:31.240 --> 0:03:34.600 the author of the Web Beneath the Waves, the fragile 0:03:34.600 --> 0:03:37.640 cables that connect our world. He's also the acting manager 0:03:37.840 --> 0:03:41.720 editor of Equator, which is a new magazine covering politics 0:03:41.720 --> 0:03:44.680 and culture. Sounds very cool, Saman, thank you so much 0:03:44.720 --> 0:03:45.920 for coming on all lots. 0:03:46.240 --> 0:03:47.120 Thank you for having me. 0:03:47.440 --> 0:03:49.080 I should have said in the intro that you're actually 0:03:49.160 --> 0:03:52.640 the author of numerous books on kind of I would say, 0:03:52.640 --> 0:03:55.720 disparate subjects. So I'm very curious why you decided to 0:03:55.760 --> 0:03:58.240 cover something like subsea cables. 0:03:58.640 --> 0:04:00.920 Well, I mean, there's a long, nonsine short answer, and 0:04:01.000 --> 0:04:03.360 the long answer has to do with this essay that 0:04:03.400 --> 0:04:06.120 I read more than ten years ago. This was an 0:04:06.200 --> 0:04:09.600 essay by the science fiction writer Neil Stephenson, and he 0:04:09.720 --> 0:04:12.640 had written this piece in Wired magazine sometime in the nineties. 0:04:12.840 --> 0:04:15.520 Forty thousand words long. It took up like virtually the 0:04:15.640 --> 0:04:21.040 entire magazine called Mother Earth, Motherboard, and that essay had Stephenson. 0:04:21.160 --> 0:04:23.000 You know, he sort of cast himself in the role 0:04:23.040 --> 0:04:25.800 of what he calls a hacker tourist, and he goes 0:04:25.800 --> 0:04:28.400 around the world kind of looking at places where these 0:04:28.440 --> 0:04:32.800 subse cables land and are installed into the earth or 0:04:32.960 --> 0:04:35.960 let into the sea, or repaired, and he meets these 0:04:36.080 --> 0:04:39.000 like odd people who do this kind of work or 0:04:39.000 --> 0:04:41.479 did this kind of work back then, and he kind 0:04:41.520 --> 0:04:44.600 of traces through all of this a picture of the 0:04:45.760 --> 0:04:48.400 nascent Internet. I mean, the Internet was already there, but 0:04:48.440 --> 0:04:50.800 I guess it wasn't as big as it was now obviously, 0:04:50.920 --> 0:04:52.680 but you know, it wasn't even as big as it 0:04:52.760 --> 0:04:54.720 was in two thousand and five. There was a lot 0:04:54.760 --> 0:04:58.440 of big data explosions yet to come, but it was 0:04:58.480 --> 0:05:02.520 clearly something that was and I think Neil Stephenson did 0:05:02.560 --> 0:05:05.480 such a good job even in the nineteen nineties, the 0:05:05.560 --> 0:05:09.960 reminding people that the Internet is reliant on this physical infrastructure, 0:05:09.960 --> 0:05:15.039 this actual cabling that looks remarkably similar to telegraph cables 0:05:15.040 --> 0:05:18.120 from the eighteen hundreds, athough of course there are big differences. 0:05:18.480 --> 0:05:20.479 So he wrote this essay in the nineteen nineties. I 0:05:20.560 --> 0:05:23.320 read it in twenty twelve, I think, and I remember 0:05:23.440 --> 0:05:27.760 being so entranced by the story that I had to 0:05:27.800 --> 0:05:30.880 step out to get groceries, and I kept reading on 0:05:30.920 --> 0:05:33.279 my phone, you know. I read in the shop, I 0:05:33.320 --> 0:05:35.960 read on the way back I was walking to and 0:05:36.000 --> 0:05:38.560 from there. And of course it struck me that the 0:05:38.560 --> 0:05:40.120 only reason I was able to load it on my 0:05:40.120 --> 0:05:43.720 phone as I was walking was because the Internet was 0:05:43.760 --> 0:05:46.240 carrying this story to me through someone to sea cable 0:05:46.320 --> 0:05:49.080 or another. And then fast forward to a few years 0:05:49.080 --> 0:05:51.880 ago when there was a big volcanic eruption off the 0:05:51.920 --> 0:05:55.440 coast of Tonga, which is an archipelago in the South Pacific, 0:05:56.320 --> 0:06:01.200 and Tonga lost connection to its only international will subseed 0:06:01.320 --> 0:06:04.960 data cable. The mudslide and landslide that kind of came 0:06:05.000 --> 0:06:08.960 out of the volcanic eruption underwater severed this line and 0:06:09.040 --> 0:06:12.320 Tongo was sort of plunged into a kind of Internet darkness. 0:06:12.800 --> 0:06:15.960 And that started to make me think, well, what is 0:06:16.000 --> 0:06:19.240 it like today for a country or a society or 0:06:19.279 --> 0:06:22.719 an economy to live without the Internet for even a 0:06:22.720 --> 0:06:25.600 brief period. What are these cables like today? How have 0:06:25.680 --> 0:06:28.400 they changed from the time that Neil Stephenson wrote this 0:06:28.520 --> 0:06:31.479 essay in the nineteen nineties. Who lays them now? Who 0:06:31.520 --> 0:06:35.039 owns them now? Has the funding changed? Do governments play 0:06:35.080 --> 0:06:38.039 a different role, similar role? And I kind of wanted 0:06:38.040 --> 0:06:39.360 to find all of this out, and I wanted to 0:06:39.440 --> 0:06:42.360 use Tonga as this kind of little test case of 0:06:42.400 --> 0:06:44.880 what it was like for a country to live without 0:06:44.880 --> 0:06:47.359 the Internet for a while. And that's why I pitched 0:06:47.360 --> 0:06:47.719 the book. 0:06:48.800 --> 0:06:52.279 So there's that famous praise from Ted Stevens, the Senator 0:06:52.320 --> 0:06:54.880 that the Internet is a series of tubes, And yeah, 0:06:54.920 --> 0:06:57.640 it is, though I never understood why everyone made fun 0:06:57.680 --> 0:06:59.720 of the guy. That's literally what we're talking about. It 0:06:59.760 --> 0:07:02.480 is a series of tubes. Okay, maybe they're like pipes 0:07:02.600 --> 0:07:05.359 or cables, but like, it never really seemed that wrong. 0:07:05.360 --> 0:07:07.520 I think the confusion comes with wireless, right where you're 0:07:07.560 --> 0:07:09.400 holding your cell phone and it's like, well, my cell 0:07:09.400 --> 0:07:12.200 phone isn't actually attached anything at the moment, so it's 0:07:12.200 --> 0:07:13.800 hard for people to wrap their heads round. 0:07:14.600 --> 0:07:17.240 I always thought you was unfairly maligned. Talk to us 0:07:17.280 --> 0:07:21.040 about the process, the very simple explanation of how a 0:07:21.080 --> 0:07:24.440 long undersea cable is laid, and how much is it 0:07:24.480 --> 0:07:28.960 fundamentally similar or different to when the first famous telegraph 0:07:29.040 --> 0:07:31.280 cable is laid between New York and London. 0:07:31.800 --> 0:07:34.400 Well, the cables themselves are very different. I mean, the 0:07:34.400 --> 0:07:37.760 first telegraph cable was made of copper and you kind 0:07:37.760 --> 0:07:40.280 of send pulses of electricity through it and that comes 0:07:40.280 --> 0:07:42.000 out of the other end and it's kind of decoded. 0:07:42.600 --> 0:07:46.320 The modern fiber optic cable is a real technological marvel. 0:07:46.360 --> 0:07:48.640 I mean, I haven't stopped marveling yet since I started 0:07:49.120 --> 0:07:51.880 researching this book. The best cables are the ones at 0:07:51.920 --> 0:07:55.080 the bottom of the ocean. Are just a hair thick, 0:07:55.520 --> 0:07:58.040 just literally the thickness of a human hair. They're made 0:07:58.080 --> 0:08:02.240 of highly purified glass and down that cable you send 0:08:02.360 --> 0:08:05.240 sort of little pulses of light lasers essentially, and they 0:08:05.280 --> 0:08:07.800 kind of bounce around the inner walls of the glass 0:08:07.840 --> 0:08:09.160 and they come out at the other end and you 0:08:09.160 --> 0:08:12.400 can decode them. And what they do these days, there's 0:08:12.440 --> 0:08:16.040 a long name for it called wave division multiplexing, where 0:08:16.040 --> 0:08:19.560 they send different frequencies of light encoded with different streams 0:08:19.600 --> 0:08:22.000 of data, and so that kind of bounces around the 0:08:22.000 --> 0:08:24.760 glass at various speeds and then it comes out of 0:08:24.760 --> 0:08:26.440 the other end. You kind of whoop, suck it out 0:08:26.480 --> 0:08:29.160 and you code it all back together and you kind 0:08:29.160 --> 0:08:31.440 of read the information out that way, and that has 0:08:31.520 --> 0:08:35.200 kind of exponentially increasy amount of data that a cable 0:08:35.240 --> 0:08:37.839 can carry to. You know, these cables are produced by 0:08:38.480 --> 0:08:41.319 just a handful of companies around the world, so there's 0:08:41.720 --> 0:08:44.319 a little bottleneck over there in terms of the technological 0:08:44.360 --> 0:08:47.920 capacity to produce these cables. And then there's people who 0:08:48.000 --> 0:08:50.200 find them money for these cables. So that used to 0:08:50.240 --> 0:08:54.600 include state owned telecom companies, then it included private investors 0:08:54.679 --> 0:08:57.480 sort of raising money from a bunch of places. Now, 0:08:57.520 --> 0:09:00.760 as we'll probably talk about, it's mostly big tech companies 0:09:00.800 --> 0:09:02.520 paying for it out of their own wallets because they 0:09:02.520 --> 0:09:05.680 can afford to. And then they would fund the a 0:09:05.760 --> 0:09:08.000 survey ship to go out and see what the best 0:09:08.120 --> 0:09:11.600 route for the cable would be from say London to 0:09:11.880 --> 0:09:15.480 Lisbon to Cape Town to I don't know, the UAE, 0:09:16.040 --> 0:09:17.719 and then the survey ship would come back and give 0:09:17.720 --> 0:09:20.160 you the best possible route, and then the cable laying 0:09:20.160 --> 0:09:23.280 ship would go out make multiple trips and just slowly, 0:09:23.559 --> 0:09:26.720 very slowly lay the cable along exactly that route. And 0:09:26.760 --> 0:09:28.679 you have to do it that way because too fast 0:09:28.679 --> 0:09:31.280 and it might snap. Too slow and there's too much 0:09:31.520 --> 0:09:33.839 slack in the cable, and so you really need to 0:09:33.880 --> 0:09:37.439 kind of find this optimum speed at which to travel. 0:09:37.720 --> 0:09:40.240 And it's really funny there's one company in the entire 0:09:40.320 --> 0:09:42.559 world that makes the kind of software that all these 0:09:42.559 --> 0:09:46.720 cable ships use that determine the ship's part, the ship speed, 0:09:46.800 --> 0:09:48.720 all of this stuff, just so the cable is laid 0:09:49.200 --> 0:09:49.480 just so. 0:09:50.320 --> 0:09:53.360 Just to be a clear quick question, is it basically 0:09:53.520 --> 0:09:56.600 whether we're talking about the Victorian at your cable or now? 0:09:57.040 --> 0:10:00.240 Is it a ship with a giant spool basically that's 0:10:00.240 --> 0:10:01.640 slowly unwine? Does it move? 0:10:02.600 --> 0:10:04.679 That's basically okay, Yeah, I mean that hasn't changed. I 0:10:04.679 --> 0:10:06.400 mean there's a couple of other things that haven't changed over, 0:10:06.440 --> 0:10:09.000 but that definitely hasn't changed. They call them a drum 0:10:09.200 --> 0:10:11.800 or they call it a spool, but essentially wind the 0:10:11.840 --> 0:10:14.480 cable around the spool and then you load it on 0:10:14.559 --> 0:10:16.600 board and then you set off into the high seas, 0:10:16.679 --> 0:10:18.960 So that part of it hasn't changed. Of course, the 0:10:18.960 --> 0:10:21.000 cable is much lighter now, so you can load a 0:10:21.040 --> 0:10:23.400 lot more cable onto it because it's not sort of 0:10:23.400 --> 0:10:24.080 thick metal. 0:10:25.360 --> 0:10:27.360 Sorry, I realized we kind of skipped ahead, but for 0:10:27.440 --> 0:10:31.959 people who who do like to malign Senator Stevens on 0:10:32.000 --> 0:10:34.840 his tube's comment, why is it that we can't just 0:10:35.000 --> 0:10:38.760 send data all through cellular means? Why do we need 0:10:38.800 --> 0:10:40.559 fiber app optic cables at all? 0:10:41.040 --> 0:10:43.200 You can, definitely, I mean, you can use satellites, and 0:10:43.240 --> 0:10:45.520 in fact, satellite phones have been around for a long while. 0:10:45.840 --> 0:10:48.520 Garmin makes all of these satellite enabled devices, and you 0:10:48.559 --> 0:10:51.840 can download weather data and other stuff onto it. It's 0:10:51.920 --> 0:10:56.120 just that the volume, the sheer volume of data that 0:10:56.240 --> 0:11:00.959 we ingest on a daily basis and fill sor a 0:11:01.000 --> 0:11:03.240 bit with enough satellites to take all that data up. 0:11:03.280 --> 0:11:06.080 I mean, think about what it means. It means everybody's 0:11:06.160 --> 0:11:09.640 Netflix streams and everybody's Zoom calls, and everybody's texts and 0:11:09.840 --> 0:11:15.280 phone calls and day trades and PowerPoint presentations that live 0:11:15.320 --> 0:11:19.040 on the cloud, and data servers that serve other kinds 0:11:19.040 --> 0:11:22.480 of things. Everything is essentially on a cloud somewhere, and 0:11:22.520 --> 0:11:25.400 there's so much data out there that there's not enough 0:11:25.440 --> 0:11:26.560 satellites that could process. 0:11:27.120 --> 0:11:29.439 Okay, that makes sense. So the other thing I really 0:11:29.480 --> 0:11:32.480 wanted to ask you is who actually makes these decisions 0:11:32.600 --> 0:11:36.439 about what gets laid and where and who's financing it? 0:11:36.559 --> 0:11:39.320 And I know you mentioned that big tech nowadays pays 0:11:39.360 --> 0:11:41.840 for most of it, But I imagine that must have 0:11:41.960 --> 0:11:44.600 changed throughout time. Right, If we're thinking back to that 0:11:44.640 --> 0:11:48.840 first Transatlantic cable, maybe it was wealthy industrialists trying to 0:11:48.880 --> 0:11:51.080 do something nice for the world. I don't know. Maybe 0:11:51.120 --> 0:11:53.520 it was governments pressing forward with it. More likely it 0:11:53.559 --> 0:11:56.880 was private companies. But the role of who's funding and 0:11:56.920 --> 0:11:59.200 planning these things must have changed throughout time. 0:12:00.040 --> 0:12:02.520 It definitely has. I mean in the eighties and nineties, 0:12:02.880 --> 0:12:05.360 and I mean I mean the nineteen eighties and nineteen nineties, 0:12:05.440 --> 0:12:07.679 a lot of these telecom companies around the world were 0:12:07.679 --> 0:12:11.040 stay owned, and so they would kind of figure out 0:12:11.040 --> 0:12:13.120 that they needed a cable to run from let's say 0:12:13.800 --> 0:12:17.439 London to Portugal, to three African countries and then onwards 0:12:17.440 --> 0:12:20.839 to Singapore as an example. So the telecom companies in 0:12:20.880 --> 0:12:23.320 each of these governments would come together and they would say, well, look, 0:12:23.360 --> 0:12:25.120 this is how we want to lay the cable. This 0:12:25.160 --> 0:12:27.560 is how much money we can put up in return, 0:12:27.640 --> 0:12:30.240 we get this much bandwidth, and we can kind of 0:12:30.280 --> 0:12:33.400 sell that onto our customers. And so these companies would 0:12:33.440 --> 0:12:35.960 form essentially like a consortium, and that was the consortium 0:12:36.000 --> 0:12:38.720 model was popular for a very long time. And then 0:12:38.840 --> 0:12:41.040 nineteen nineties in the early two thousands, you saw this 0:12:41.200 --> 0:12:44.640 wave of privatization around the world. All of these telecom 0:12:44.720 --> 0:12:49.719 companies were hived off into private companies. And suddenly at 0:12:49.760 --> 0:12:52.440 the same time there was also this growth of the 0:12:52.480 --> 0:12:55.360 investor led model. Like an investor would say, this cable 0:12:55.400 --> 0:12:57.920 looks like a good business idea, I will go out 0:12:57.960 --> 0:13:00.440 and raise money. And then I was come back to 0:13:00.440 --> 0:13:02.439 these telecom companies and say, look, I'm willing to lay 0:13:02.440 --> 0:13:04.560 this cable. How much bandwidth do you want to buy 0:13:04.600 --> 0:13:07.840 from me? And they would kind of apportion bandwidth in 0:13:07.880 --> 0:13:10.640 that way. And so that model prevailed for like much 0:13:10.679 --> 0:13:14.000 of the two thousands. And I think it's only about 0:13:14.600 --> 0:13:18.640 seven or eight years ago that big tech companies, by 0:13:18.640 --> 0:13:23.800 which I mean Google, Meta, Amazon, and Microsoft A grew 0:13:23.840 --> 0:13:26.720 so cash rich that they could afford these cables. So, 0:13:26.800 --> 0:13:30.080 for example, a transatlantic cable now from London to New 0:13:30.160 --> 0:13:33.280 York would cost about five hundred million dollars, which there's 0:13:33.320 --> 0:13:35.320 a lot of money still for you and me, but 0:13:35.400 --> 0:13:37.800 maybe not that much for Google. And so they started 0:13:37.800 --> 0:13:40.800 to fund these cables because rightly or wrongly, they thought, look, 0:13:40.920 --> 0:13:42.960 data is the lifeblood of our business, and it makes 0:13:43.000 --> 0:13:46.360 more sense to build this infrastructure ourselves. And so now 0:13:46.400 --> 0:13:48.319 we got to the stage, or when I was researching 0:13:48.360 --> 0:13:51.040 this book, it got to a stage where two out 0:13:51.080 --> 0:13:55.280 of every three new cables were being funded and owned, 0:13:55.400 --> 0:13:57.400 either in part or in full, by one of these 0:13:57.400 --> 0:14:02.040 four tech companies. And that has enormous sort of implications 0:14:02.080 --> 0:14:06.000 for data privacy, data security. I mean rolls your data 0:14:06.160 --> 0:14:08.599 also gets the Internet. 0:14:23.960 --> 0:14:26.320 When the infrastructure of the Internet was first starting to 0:14:26.320 --> 0:14:29.120 be built out, particularly in the nineteen nineties. There's a 0:14:29.160 --> 0:14:32.960 popular parallel to the railway booms, et cetera. And it's 0:14:32.960 --> 0:14:36.760 sort of similar economics and connecting locations and all. That 0:14:37.360 --> 0:14:40.720 rail is kind of a natural monopoly. You know, if 0:14:40.760 --> 0:14:43.680 you want to have a rail between New York and Chicago, 0:14:44.160 --> 0:14:48.080 you don't need five separate tracks owned by separate companies, 0:14:48.120 --> 0:14:50.600 all right next to each other. It's sort of wasteful, 0:14:50.800 --> 0:14:51.280 et cetera. 0:14:51.440 --> 0:14:53.960 Wait, let me just say, as a regular Amtrak commuter, 0:14:54.080 --> 0:14:55.240 I do wish there were. 0:14:55.200 --> 0:14:58.400 Multiple wish there were competition, but there's never gonna be right. 0:14:58.440 --> 0:15:00.480 That'll just never happen. There aren't going to be four 0:15:00.640 --> 0:15:03.200 tracks between a random town and Connecticut, and you're what 0:15:03.320 --> 0:15:06.560 one can dream? One can dream? To what degree do 0:15:06.680 --> 0:15:11.000 the economics of this replicate, how much redundancy is there, 0:15:11.200 --> 0:15:15.200 how many different cables are there, in some cases along 0:15:15.560 --> 0:15:19.240 exactly the same route. To what degree does the natural 0:15:19.240 --> 0:15:21.120 monopoly effect either replicated or not. 0:15:22.440 --> 0:15:25.640 Well, there's a considerable amount of redundancy built in because 0:15:25.640 --> 0:15:29.160 I think the deal here is this, even if Google owns, 0:15:29.200 --> 0:15:33.000 for example, one transatlantic cable, and even if it, let's 0:15:33.000 --> 0:15:37.920 assume hypothetically, it funnels only Google based data through that cable. 0:15:37.960 --> 0:15:40.880 So let's assume funnels Gmail and Google Meets and all 0:15:40.920 --> 0:15:43.720 this other stuff. Only Google data is going through that cable. 0:15:43.840 --> 0:15:46.720 Is this never happens, but we'll let it ride from now, 0:15:47.560 --> 0:15:52.040 It is in Google's best interests to buy redundancy on 0:15:52.080 --> 0:15:54.200 another cable because that cable might go down. And in 0:15:55.120 --> 0:15:57.760 a similar way, it is in Meta's best interest to 0:15:57.800 --> 0:16:00.240 buy redundancy on a Google cable just in case Meta 0:16:00.280 --> 0:16:02.600 a cable goes down. So there's a lot of redundancy 0:16:02.600 --> 0:16:05.600 built into the system. There's between five hundred and five 0:16:05.640 --> 0:16:09.520 hundred and fifty undersea cables around the world, many of 0:16:09.560 --> 0:16:13.560 them are sort of clustered around these not surprisingly these 0:16:13.600 --> 0:16:16.880 big areas of economic activity, so Western Europe to the 0:16:16.880 --> 0:16:21.080 eastern seaboard of the US, Southeast Asia, China and Southeast Asia, 0:16:21.200 --> 0:16:24.760 the Gulf, and then India. So these are heavily traffic 0:16:25.440 --> 0:16:27.440 routes and there's a lot of cables that traffic them 0:16:27.760 --> 0:16:30.760 for the simple reason that redundancy is important. We can 0:16:30.800 --> 0:16:35.120 think about the fact that every year, roughly one hundred 0:16:35.440 --> 0:16:39.119 cables get cut around the world. Most of these are accidents. 0:16:39.400 --> 0:16:42.200 There might be a ship that throws its anchor overboard 0:16:42.280 --> 0:16:44.160 and cuts a cable by accident, or it might be 0:16:44.200 --> 0:16:46.960 that a fishing boat is trawling the seabed and it 0:16:47.000 --> 0:16:49.520 hooks a cable and snags it and cuts it. So 0:16:49.600 --> 0:16:51.720 think about one hundred cuts every year, and yet we 0:16:51.800 --> 0:16:57.560 don't experience nearly that order of internet outages, right, And 0:16:57.600 --> 0:16:59.360 part of the reason is, of course, we have redundancy 0:16:59.360 --> 0:17:01.360 in the form of le cables, which I haven't really 0:17:01.360 --> 0:17:04.439 talked about in this book, but there's cables coming to 0:17:04.480 --> 0:17:07.120 the US from Canada and so on, so that takes 0:17:07.160 --> 0:17:09.760 up a share of the traffic. But also there's redundancy 0:17:09.840 --> 0:17:12.800 in the marine cable ecosystem itself, so like if one 0:17:12.880 --> 0:17:17.000 cable gets cut, there's enough cables crossing the Atlantic, for example, 0:17:17.400 --> 0:17:20.320 to take up the burden of the data that was severed. 0:17:21.640 --> 0:17:23.440 Actually, this reminds me the other thing I wanted to 0:17:23.480 --> 0:17:26.720 ask you is how much does geography play into these decisions? 0:17:26.760 --> 0:17:29.840 Because again going back to that chart, they very much 0:17:29.880 --> 0:17:33.840 resemble shipping lanes, so getting from point A to point 0:17:33.840 --> 0:17:37.840 B presumably as fast and efficiently as possible. But also 0:17:37.960 --> 0:17:41.040 one thing I learned in researching for this episode is 0:17:41.080 --> 0:17:44.920 that Egypt apparently is a big I don't want to 0:17:44.920 --> 0:17:48.760 say choke point, but like a major center for fiber 0:17:48.800 --> 0:17:51.399 optic cables in much the same way that it's a 0:17:51.680 --> 0:17:55.320 major center for containers shipping through the Suez Canal. So 0:17:55.640 --> 0:17:58.720 how much does just pure basic geography actually inform the 0:17:58.720 --> 0:18:01.359 decisions of where these cables get laid. 0:18:02.240 --> 0:18:04.840 Oh, it's a big deal. I mean first, there's under 0:18:05.080 --> 0:18:07.639 sea geography, right, I mean, the topography of the ocean 0:18:07.640 --> 0:18:11.199 floor is obviously not flat flat flat. There are crevasses 0:18:11.200 --> 0:18:13.920 and gorgeos, and there's currents that go one way or another. 0:18:13.960 --> 0:18:18.119 There's like steep lips of rocky cliffs, and so you 0:18:18.160 --> 0:18:21.120 have to kind of avoid all of this. You know, Tonga, 0:18:21.160 --> 0:18:24.960 which I mentioned earlier, has the unfortunate situation of sitting 0:18:24.960 --> 0:18:27.080 in that Pacific Ring of Fire. So there's an underwater 0:18:27.160 --> 0:18:30.879 volcano not far away and that's liable to break its cables, 0:18:30.880 --> 0:18:32.680 and it has in the past, even before the one 0:18:32.680 --> 0:18:36.120 that I talk about. So under sea geology geography big deal. 0:18:36.560 --> 0:18:39.560 But even overseas geography is a massive deal. So for example, 0:18:39.560 --> 0:18:42.240 if you want to come from Asia to Europe as 0:18:42.240 --> 0:18:44.639 with a ship, it's much cheaper to come through this 0:18:44.760 --> 0:18:48.120 West Canal rather than to go around around South Africa. 0:18:48.560 --> 0:18:51.960 And you know, in the Neil Stephenson essay I talked about, 0:18:52.000 --> 0:18:53.320 he makes a big deal out of this is a 0:18:53.560 --> 0:18:57.560 really elegant way. He goes to Alexandria, where a bunch 0:18:57.600 --> 0:19:00.440 of cables are sort of crisscrossing over from Asia to 0:19:00.520 --> 0:19:03.440 Europe and vice versa, and he talks about how Alexandria 0:19:03.520 --> 0:19:06.320 is at that point kind of like a storehouse or 0:19:06.760 --> 0:19:09.399 a hub of information, very similar to how it was 0:19:09.800 --> 0:19:11.680 back in the day when the Library of Alexandria was 0:19:11.720 --> 0:19:15.280 still in place. So there's this parallel with ancient history. 0:19:16.000 --> 0:19:18.439 But you know, Egypt, you bring up Egypt, and Egypt 0:19:18.440 --> 0:19:21.560 is a really good example. The other good example is 0:19:21.560 --> 0:19:23.280 something that's very much in the news now, which is 0:19:23.280 --> 0:19:27.120 the Strait of Hormons. And just as it's very convenient 0:19:27.160 --> 0:19:29.600 for ships to take that path in, so it is 0:19:29.680 --> 0:19:31.680 for cables to take that path in. And the Red 0:19:31.720 --> 0:19:34.280 Sea and the Strait of Hormoves are both chock points, 0:19:34.280 --> 0:19:37.040 as you say again, in which if a malign actors 0:19:37.040 --> 0:19:39.639 sort of wanted to really cripple the global Internet, they 0:19:39.680 --> 0:19:42.119 could go down there, use your lobster drones that you 0:19:42.200 --> 0:19:45.280 talked about, and clip every single cable that runs through 0:19:45.320 --> 0:19:46.400 the ocean floor over there. 0:19:46.920 --> 0:19:48.840 Wait, can I just ask on that? So when I 0:19:48.880 --> 0:19:52.199 look at the map of subc cables going across the 0:19:52.200 --> 0:19:55.159 Strait of Horror Moves, it's like Saudi Arabia connecting to 0:19:55.560 --> 0:19:59.360 I guess Iran. Can they cripple the global Internet by 0:20:00.040 --> 0:20:03.240 apping those cables in particular, or would it just cripple 0:20:03.280 --> 0:20:06.280 the connectivity between the countries that are clustered around there. 0:20:06.880 --> 0:20:10.560 Well, I mean it would definitely impede the countries trusted 0:20:10.600 --> 0:20:12.119 around there. But we also have to remember that the 0:20:12.160 --> 0:20:15.760 Internet doesn't function like a highway in the sense that 0:20:16.200 --> 0:20:19.040 the fastest route between two points is not necessarily the 0:20:19.760 --> 0:20:22.119 shortest route. Okay. So for example, if I'm sitting in 0:20:22.160 --> 0:20:26.720 London and I'm sort of pinging a server in let's 0:20:26.760 --> 0:20:30.399