1 00:00:10,560 --> 00:00:12,640 Speaker 1: We have breaking news for you this morning. 2 00:00:13,119 --> 00:00:17,279 Speaker 2: A widespread Internet outage has caused major disruptions for some 3 00:00:17,360 --> 00:00:18,919 Speaker 2: of the biggest web services in the world. 4 00:00:18,960 --> 00:00:21,880 Speaker 3: There's a good chance today that some website, app or 5 00:00:21,920 --> 00:00:23,919 Speaker 3: service that you tried to use today was down. That's 6 00:00:23,920 --> 00:00:26,120 Speaker 3: following an outage from Amazon Web Services. 7 00:00:26,160 --> 00:00:29,800 Speaker 4: Amazon Web Services, the backbone of the Internet, suffered a 8 00:00:29,880 --> 00:00:34,520 Speaker 4: major outage that took hundreds of services offline across the globe. 9 00:00:34,560 --> 00:00:37,960 Speaker 4: Monday morning, October twentieth, Americans woke up to news of 10 00:00:38,040 --> 00:00:41,320 Speaker 4: a massive Internet outage, and it wasn't just America. 11 00:00:41,840 --> 00:00:44,199 Speaker 1: I think one of the things to really note about 12 00:00:44,240 --> 00:00:47,600 Speaker 1: this outage is that it was one of the biggest 13 00:00:47,600 --> 00:00:52,960 Speaker 1: in recent history. In my decade in the field, this 14 00:00:53,360 --> 00:00:55,200 Speaker 1: definitely like ranks in the top five. 15 00:00:56,000 --> 00:00:58,800 Speaker 4: Doctor Kreen caf is a culture anthropologist and she works 16 00:00:58,800 --> 00:01:03,840 Speaker 4: at the Human Rights organization Article nineteen. She researches Internet infrastructure, AI, 17 00:01:04,040 --> 00:01:07,480 Speaker 4: cloud computing, and basically how technology affects our day to 18 00:01:07,560 --> 00:01:08,360 Speaker 4: day lives. 19 00:01:08,680 --> 00:01:11,680 Speaker 1: I have asked some of my colleagues, like veterans of 20 00:01:11,720 --> 00:01:14,080 Speaker 1: the field who've been in the industry since the Internet 21 00:01:14,160 --> 00:01:17,120 Speaker 1: was born, and they pretty much said the same thing, 22 00:01:17,240 --> 00:01:20,160 Speaker 1: like this was huge a. 23 00:01:20,120 --> 00:01:22,160 Speaker 4: Lot of people didn't necessarily know that there was some 24 00:01:22,200 --> 00:01:24,880 Speaker 4: global internet outage going on. They just knew that there 25 00:01:24,959 --> 00:01:27,640 Speaker 4: was some website or service that they used that wasn't working, 26 00:01:28,000 --> 00:01:31,319 Speaker 4: like their favorite messaging app or even their smart bed. 27 00:01:32,360 --> 00:01:35,959 Speaker 1: Some of these are kind of beIN benign or even funny, 28 00:01:36,040 --> 00:01:39,240 Speaker 1: like there were some excellent threads on the Internet about 29 00:01:39,680 --> 00:01:43,240 Speaker 1: people who had Internet connected mattresses. 30 00:01:44,160 --> 00:01:45,240 Speaker 4: Oh the beds. 31 00:01:45,600 --> 00:01:49,000 Speaker 2: Yes, yes, the beds. Oh my god, Apods. Shout out 32 00:01:49,040 --> 00:01:50,720 Speaker 2: to apods and everyone who thought it was a good 33 00:01:50,720 --> 00:01:52,840 Speaker 2: idea to buy an internet connected bed. I sort of 34 00:01:52,880 --> 00:01:55,120 Speaker 2: got this is why we don't have a single interconnected 35 00:01:55,240 --> 00:01:56,720 Speaker 2: device in our house. 36 00:01:57,160 --> 00:01:59,840 Speaker 4: Krin's talking about these really expensive smart beds by this 37 00:02:00,040 --> 00:02:03,520 Speaker 4: company called eight Sleep. Their mattresses are supposed to track 38 00:02:03,520 --> 00:02:06,240 Speaker 4: your biometric data and get hotter or cooler, depending on 39 00:02:06,280 --> 00:02:08,960 Speaker 4: what your body needs. They're also connected to the Internet, 40 00:02:09,080 --> 00:02:13,080 Speaker 4: and apparently they didn't work offline after the outage started. 41 00:02:13,240 --> 00:02:17,280 Speaker 4: Some of the mattresses overheated or wouldn't recline. Thinking about 42 00:02:17,320 --> 00:02:21,200 Speaker 4: somebody's twenty five hundred dollars internet smartbed going vertical and 43 00:02:21,280 --> 00:02:24,600 Speaker 4: trying to cook them alive is kind of dystopian, but 44 00:02:24,880 --> 00:02:27,639 Speaker 4: it goes beyond that. This thing took down everything from 45 00:02:27,639 --> 00:02:31,320 Speaker 4: Amazon's own products like Ring doorbells and Alexa, to sites 46 00:02:31,360 --> 00:02:35,120 Speaker 4: like Reddit, to messaging apps like Signal, to government services. 47 00:02:35,960 --> 00:02:39,080 Speaker 4: These outages can impact basically every part of your life. 48 00:02:39,720 --> 00:02:43,360 Speaker 1: This isn't just like me being on Duo Lingo or 49 00:02:43,440 --> 00:02:49,279 Speaker 1: me being on Snapchat. It is banking, It is the hospitals, 50 00:02:49,320 --> 00:02:51,920 Speaker 1: it is education, like it is things that matter in 51 00:02:51,919 --> 00:02:54,200 Speaker 1: my life, that matter in everyone's life. 52 00:02:57,080 --> 00:03:00,920 Speaker 4: The world today relies on basically three major cloud comuting services. 53 00:03:01,160 --> 00:03:03,360 Speaker 4: When one of those goes down, so does a huge 54 00:03:03,440 --> 00:03:06,960 Speaker 4: chunk of the Internet. So what causes these outages? How 55 00:03:06,960 --> 00:03:08,560 Speaker 4: do we get to the point where so much of 56 00:03:08,600 --> 00:03:11,840 Speaker 4: the Internet relies on just three big tech companies? And 57 00:03:12,040 --> 00:03:25,639 Speaker 4: can that ever change any time? From Kaleidoscope and iHeart Podcast, 58 00:03:26,200 --> 00:03:29,240 Speaker 4: this is kill Switch. I'm Dexter Thomas. 59 00:03:37,320 --> 00:03:40,320 Speaker 1: Goodbye. 60 00:03:43,280 --> 00:03:47,240 Speaker 4: What exactly happened on October twentieth, So. 61 00:03:47,400 --> 00:03:51,320 Speaker 1: What essentially happened is that there was an outage within 62 00:03:51,400 --> 00:03:55,480 Speaker 1: Amazon web Services. Now, Amazon Web Services is a leading 63 00:03:55,680 --> 00:03:58,760 Speaker 1: cloud computing company, and what is super important to know 64 00:03:59,040 --> 00:04:02,000 Speaker 1: is that Amazon currently has at least thirty percent of 65 00:04:02,040 --> 00:04:03,640 Speaker 1: the global cloud market share. 66 00:04:05,680 --> 00:04:09,080 Speaker 4: We sometimes think of the cloud as something that lives 67 00:04:09,080 --> 00:04:12,600 Speaker 4: in this virtual space, not as something tangible. But the 68 00:04:12,600 --> 00:04:16,159 Speaker 4: cloud relies on very real and physical data centers that 69 00:04:16,160 --> 00:04:19,120 Speaker 4: are located around the world. These things store code and 70 00:04:19,279 --> 00:04:21,880 Speaker 4: data and move around that data to make the apps 71 00:04:21,920 --> 00:04:25,599 Speaker 4: that you use or the Internet run. And as Caring 72 00:04:25,720 --> 00:04:28,239 Speaker 4: just pointed out, about a third of the global Internet 73 00:04:28,279 --> 00:04:32,400 Speaker 4: relies on one company for cloud computing, Amazon Web Services. 74 00:04:33,320 --> 00:04:37,440 Speaker 1: This company operates data centers worldwide that a lot of 75 00:04:37,440 --> 00:04:41,440 Speaker 1: the Internet and a lot of applications rely on. And 76 00:04:41,480 --> 00:04:45,440 Speaker 1: so what happened is within their data center clusters, one 77 00:04:45,480 --> 00:04:49,840 Speaker 1: of the oldest and biggest one had a bug in 78 00:04:49,880 --> 00:04:54,240 Speaker 1: the automation software that subsequently took down all. 79 00:04:54,120 --> 00:04:56,440 Speaker 3: Of the companies that relied on that data center. 80 00:04:56,880 --> 00:04:59,880 Speaker 4: Basically, the reason this outage was so widespread was because 81 00:05:00,000 --> 00:05:02,839 Speaker 4: it was an issue with one specific and heavily relied 82 00:05:02,920 --> 00:05:06,320 Speaker 4: upon data center that's located in Virginia. 83 00:05:06,520 --> 00:05:10,360 Speaker 1: I've seen estimates of you know, a billion pins per 84 00:05:10,480 --> 00:05:13,320 Speaker 1: day going to this particular data center in Virginia, and 85 00:05:13,360 --> 00:05:15,520 Speaker 1: a pin is essentially like when a company needs something 86 00:05:15,560 --> 00:05:18,560 Speaker 1: from that database. And because this data center was one 87 00:05:18,600 --> 00:05:21,760 Speaker 1: of the biggest, some one of the oldest, about thirty 88 00:05:21,800 --> 00:05:25,360 Speaker 1: five to forty percent of all AWS traffic. 89 00:05:25,520 --> 00:05:26,240 Speaker 3: Goes through it. 90 00:05:27,920 --> 00:05:32,279 Speaker 1: Another reason why this impact was so big is if 91 00:05:32,320 --> 00:05:35,800 Speaker 1: you are a software developer and like you buy AWS 92 00:05:35,839 --> 00:05:39,479 Speaker 1: to develop your software on, this region is the default, 93 00:05:40,240 --> 00:05:43,520 Speaker 1: So like, if you don't switch to the default, this 94 00:05:43,560 --> 00:05:44,919 Speaker 1: is where your stuff will be housed. 95 00:05:46,080 --> 00:05:48,520 Speaker 4: And so you have a third of the world's Internet 96 00:05:48,600 --> 00:05:52,279 Speaker 4: relying on this one company. Within that, there's this one 97 00:05:52,400 --> 00:05:54,839 Speaker 4: data center that becomes a default for all kinds of 98 00:05:54,839 --> 00:05:58,159 Speaker 4: new apps and websites. And then last month that data 99 00:05:58,160 --> 00:06:02,039 Speaker 4: center had a DNS error. The DNS, the domain name system, 100 00:06:02,320 --> 00:06:05,080 Speaker 4: translates a written website into an IP address that a 101 00:06:05,120 --> 00:06:07,839 Speaker 4: computer can then decode and understand. You might have heard 102 00:06:07,839 --> 00:06:10,600 Speaker 4: this described as the phone book of the Internet, but 103 00:06:10,800 --> 00:06:12,360 Speaker 4: Carin has a better explanation. 104 00:06:13,080 --> 00:06:14,159 Speaker 3: So I'm going to. 105 00:06:14,200 --> 00:06:17,760 Speaker 1: Try and do a different metaphor to explain this, because 106 00:06:18,640 --> 00:06:20,720 Speaker 1: the phone book is really useful in a sense that 107 00:06:20,760 --> 00:06:24,479 Speaker 1: what it does it connects names to numbers, right, and 108 00:06:24,520 --> 00:06:27,359 Speaker 1: that is what the domain name system does. So when 109 00:06:27,520 --> 00:06:31,640 Speaker 1: you type in a websites, you type in riverside dot FM, 110 00:06:32,080 --> 00:06:35,200 Speaker 1: not the very long number that is behind it, which 111 00:06:35,240 --> 00:06:39,600 Speaker 1: is actually how machines find each other. The problem with 112 00:06:39,720 --> 00:06:43,720 Speaker 1: the phone book metaphor is I've held one, I know 113 00:06:43,800 --> 00:06:45,720 Speaker 1: what they look like and what they feel like. But 114 00:06:46,000 --> 00:06:49,000 Speaker 1: most people younger than me, that is not a helpful 115 00:06:49,040 --> 00:06:51,800 Speaker 1: metaphor at all. So the way that I try to 116 00:06:51,800 --> 00:06:56,920 Speaker 1: explain what happened with AWS is imagine you're using Google 117 00:06:56,960 --> 00:07:00,520 Speaker 1: Maps or Ways or some sort of navigation app because 118 00:07:00,560 --> 00:07:02,280 Speaker 1: you want to go to the library because you need 119 00:07:02,320 --> 00:07:05,479 Speaker 1: a specific book to do your job. But essentially what 120 00:07:05,520 --> 00:07:09,239 Speaker 1: happens is there's a software bug and that causes your app, 121 00:07:09,320 --> 00:07:12,560 Speaker 1: your map app, to essentially indicate that the library no 122 00:07:12,600 --> 00:07:15,480 Speaker 1: longer exists. Even though the library is right there, it 123 00:07:15,520 --> 00:07:18,560 Speaker 1: is fully operational, but you just no longer can see 124 00:07:18,560 --> 00:07:21,040 Speaker 1: how to get there, and you're stuck with two problems, 125 00:07:21,080 --> 00:07:25,040 Speaker 1: the first one being is you cannot navigate to the library, 126 00:07:25,120 --> 00:07:28,920 Speaker 1: and secondly, you cannot access the information that you need, 127 00:07:29,000 --> 00:07:31,960 Speaker 1: and so you can't do your work. And that's essentially 128 00:07:32,000 --> 00:07:35,480 Speaker 1: what happened with the AWS outage. But the problem is, 129 00:07:35,480 --> 00:07:37,680 Speaker 1: like you're not the only one who needs the library. 130 00:07:38,200 --> 00:07:42,400 Speaker 1: The writer on your team needs access to it for research, 131 00:07:42,520 --> 00:07:45,880 Speaker 1: the accountant needs it for whatever financial records or software, 132 00:07:46,200 --> 00:07:49,440 Speaker 1: your lawyer needs it for the database. With case law, right, 133 00:07:49,520 --> 00:07:52,560 Speaker 1: But when the map says the library doesn't exist, it 134 00:07:52,680 --> 00:07:56,880 Speaker 1: pretty much doesn't show up, and so everyone's work stops simultaneously. 135 00:07:57,520 --> 00:08:01,119 Speaker 1: And what AWS figured out is like fixing that map 136 00:08:01,240 --> 00:08:05,280 Speaker 1: is not easy. You have to take all of these steps. 137 00:08:05,280 --> 00:08:07,800 Speaker 1: You have to identify which version of the mapping software 138 00:08:07,800 --> 00:08:10,680 Speaker 1: each person is using, push all of the updates to 139 00:08:10,760 --> 00:08:13,400 Speaker 1: all of the systems, wait for everything to restart, and 140 00:08:13,440 --> 00:08:16,080 Speaker 1: then verify that everyone can see the library again on 141 00:08:16,120 --> 00:08:18,520 Speaker 1: the map. And then you get this really interesting part 142 00:08:18,560 --> 00:08:21,760 Speaker 1: where when you fixed it and the library is reachable again, 143 00:08:22,240 --> 00:08:26,160 Speaker 1: there's a stampede because everyone realizes, like, oh, it's back up, 144 00:08:26,200 --> 00:08:28,000 Speaker 1: so we now all need to go there. And everyone 145 00:08:28,040 --> 00:08:32,360 Speaker 1: who's done like Black Friday shopping knows what that leads to. 146 00:08:32,840 --> 00:08:34,760 Speaker 1: And that is also why it took so long, right like, 147 00:08:34,800 --> 00:08:37,920 Speaker 1: because one thing is fixing the problem, and subsequently you 148 00:08:38,000 --> 00:08:40,480 Speaker 1: have to deal with the backlog of everyone trying to 149 00:08:40,640 --> 00:08:43,040 Speaker 1: get through the door simultaneously and the system not being 150 00:08:43,080 --> 00:08:44,120 Speaker 1: able to deal with that either. 151 00:08:44,960 --> 00:08:47,200 Speaker 4: This isn't the first time the cloud services have had 152 00:08:47,200 --> 00:08:50,079 Speaker 4: an error like this, but because of how many applications 153 00:08:50,080 --> 00:08:53,439 Speaker 4: rely not only on AWS, but on this one data center, 154 00:08:53,800 --> 00:08:57,280 Speaker 4: the effects of the outage were really widespread, and outages 155 00:08:57,320 --> 00:09:00,000 Speaker 4: of this scale can cause some very real problem. 156 00:09:00,880 --> 00:09:03,080 Speaker 1: This is not the first outage, not the first and 157 00:09:03,080 --> 00:09:05,520 Speaker 1: big one, right. The last big one was on the 158 00:09:05,600 --> 00:09:08,840 Speaker 1: nineteenth of July twenty twenty four, And the reason why 159 00:09:08,920 --> 00:09:12,880 Speaker 1: I remember that date so precisely is because it happens 160 00:09:12,920 --> 00:09:15,199 Speaker 1: to be the day that my second born was born, 161 00:09:15,600 --> 00:09:18,960 Speaker 1: and for reasons I had to be rushed to the 162 00:09:19,080 --> 00:09:20,920 Speaker 1: hospital and it was like urgent. 163 00:09:20,960 --> 00:09:21,640 Speaker 3: This was no joke. 164 00:09:22,080 --> 00:09:24,640 Speaker 1: And that was another day that there was a massive outage, 165 00:09:24,640 --> 00:09:27,040 Speaker 1: but this time it was CrowdStrike Microsoft Cloud. 166 00:09:27,920 --> 00:09:30,400 Speaker 4: In this incident last year in twenty twenty four, what 167 00:09:30,520 --> 00:09:33,600 Speaker 4: happened was a faulty update by the security company CrowdStrike 168 00:09:33,880 --> 00:09:37,640 Speaker 4: caused a major outage for Microsoft's Azure cloud services. This 169 00:09:37,679 --> 00:09:41,599 Speaker 4: one brought down airlines, banks, and even emergency services. 170 00:09:42,000 --> 00:09:46,280 Speaker 1: The massive Microsoft outage impacting businesses around the world, taking 171 00:09:46,320 --> 00:09:50,760 Speaker 1: some companies temporarily offline and forcing major American airlines to 172 00:09:50,880 --> 00:09:55,400 Speaker 1: ground flights. The outage affected government entities and news organizations, as. 173 00:09:55,360 --> 00:09:59,839 Speaker 4: Well as almost every industry across the globe, including banks, airports, 174 00:10:00,040 --> 00:10:01,920 Speaker 4: healthcare providers, and retailers. 175 00:10:03,320 --> 00:10:06,719 Speaker 1: There were a number of hospitals in the Netherlands where 176 00:10:06,760 --> 00:10:10,320 Speaker 1: I'm based, where those issues were such that they couldn't 177 00:10:10,320 --> 00:10:13,560 Speaker 1: help patients. Now luckily that wasn't in my region. But 178 00:10:14,240 --> 00:10:18,160 Speaker 1: when you're in an ambulance being sped to a hospital, you 179 00:10:18,280 --> 00:10:20,000 Speaker 1: don't want to show up there on the other end 180 00:10:20,040 --> 00:10:23,960 Speaker 1: and then telling you I'm sorry, computer says no. You know, 181 00:10:24,000 --> 00:10:26,679 Speaker 1: it's the apps and it's the benign stuff, but it's 182 00:10:26,720 --> 00:10:29,240 Speaker 1: also the very serious stuff that we have connected to 183 00:10:29,280 --> 00:10:35,520 Speaker 1: these clouds. Healthcare, education, banking, access to your government. These 184 00:10:35,559 --> 00:10:39,320 Speaker 1: are all things that are so critical for people's lives. 185 00:10:39,679 --> 00:10:42,520 Speaker 1: Like what happens when that is not accessible? I mean, 186 00:10:42,600 --> 00:10:45,439 Speaker 1: I am sure that there are people who missed funerals, 187 00:10:45,760 --> 00:10:48,480 Speaker 1: who were not able to make really crucial payments that 188 00:10:48,640 --> 00:10:51,520 Speaker 1: needed to be made that day, that didn't get whatever 189 00:10:51,640 --> 00:10:53,880 Speaker 1: benefits that they really needed to do. This is not 190 00:10:54,120 --> 00:10:58,120 Speaker 1: an incident because it keeps happening. When this happened, like 191 00:10:58,320 --> 00:11:01,400 Speaker 1: the organization that I work for nineteen like, we had 192 00:11:01,400 --> 00:11:05,960 Speaker 1: a blogout within two hours at best, and people were like, wow, 193 00:11:06,000 --> 00:11:08,040 Speaker 1: you guys are so fast, like how did you put 194 00:11:08,040 --> 00:11:10,079 Speaker 1: that together? And I'm like, well, it's kind of depressing 195 00:11:10,120 --> 00:11:13,240 Speaker 1: because this is the fourth time in four years that 196 00:11:13,280 --> 00:11:15,920 Speaker 1: I see in outage quite like this, And all I 197 00:11:16,000 --> 00:11:17,880 Speaker 1: have to do every time is I switch out the 198 00:11:17,960 --> 00:11:21,120 Speaker 1: name of the company and I look into the specifics 199 00:11:21,360 --> 00:11:24,920 Speaker 1: of where they feel technically, and then it's the same 200 00:11:25,000 --> 00:11:28,559 Speaker 1: up D and so we just you know, here we are. 201 00:11:29,160 --> 00:11:32,560 Speaker 4: That is depressing. That is depressing. It's wild too, you 202 00:11:32,600 --> 00:11:35,400 Speaker 4: said it keeps happening. You might already know this, but 203 00:11:35,760 --> 00:11:39,720 Speaker 4: I asked you on to talk about aws. Do you 204 00:11:39,760 --> 00:11:41,240 Speaker 4: know Minecraft is down right now? 205 00:11:41,480 --> 00:11:41,680 Speaker 2: Oh? 206 00:11:41,760 --> 00:11:42,600 Speaker 3: I did not see that. 207 00:11:43,200 --> 00:11:47,400 Speaker 1: Yeah, preparing for talking to you guys and not. 208 00:11:46,320 --> 00:11:49,040 Speaker 4: Not doing your job, I kept you from doing your job. Yeah, No, 209 00:11:49,559 --> 00:11:54,320 Speaker 4: I think Xbox is down right now. Microsoft's Azure was 210 00:11:54,360 --> 00:11:57,480 Speaker 4: once again experiencing an outage. At that point in the interview, 211 00:11:57,520 --> 00:11:59,960 Speaker 4: all I'd seen was that Xbox and Minecraft were down. 212 00:12:00,280 --> 00:12:03,160 Speaker 4: But as always, it's bigger than that. It turns out 213 00:12:03,240 --> 00:12:05,520 Speaker 4: banks got hit with a two and so did parts 214 00:12:05,520 --> 00:12:08,440 Speaker 4: of the train system, which was a problem because it 215 00:12:08,520 --> 00:12:11,000 Speaker 4: also happened to be election day in the Netherlands, where 216 00:12:11,080 --> 00:12:13,800 Speaker 4: Krin is, and the disruptions and Dutch railways meant that 217 00:12:13,880 --> 00:12:15,760 Speaker 4: some people might not be able to make at home 218 00:12:15,760 --> 00:12:19,359 Speaker 4: in time to vote. So the outage was directly interfering 219 00:12:19,360 --> 00:12:23,160 Speaker 4: with the democratic process. Microsoft's share of cloud computing isn't 220 00:12:23,160 --> 00:12:25,920 Speaker 4: as big as Amazon's, but it's a close second, and 221 00:12:25,960 --> 00:12:29,600 Speaker 4: together with Google, these three companies make up almost two 222 00:12:29,720 --> 00:12:33,680 Speaker 4: thirds of the entire cloud infrastructure, meaning that two thirds 223 00:12:33,720 --> 00:12:38,320 Speaker 4: of the Internet relies on Google, Amazon, and Microsoft keeping 224 00:12:38,360 --> 00:12:42,400 Speaker 4: their act together. So how did we get here after 225 00:12:42,440 --> 00:12:54,079 Speaker 4: the break the origin story of Amazon Web Services? How 226 00:12:54,080 --> 00:12:56,320 Speaker 4: do we get to the point where the bookseller became 227 00:12:56,840 --> 00:12:58,920 Speaker 4: the backbone of most of the Internet. 228 00:12:59,080 --> 00:13:03,280 Speaker 1: So there are more multiple origin stories around, like how 229 00:13:03,320 --> 00:13:06,640 Speaker 1: AWS became AWS. A lot of the work that people 230 00:13:06,720 --> 00:13:11,480 Speaker 1: put out there now like very much follows aws's iteration 231 00:13:11,600 --> 00:13:15,720 Speaker 1: of events. You know, we were an e commerce platform. 232 00:13:16,040 --> 00:13:19,200 Speaker 1: We needed a lot of compute to be able to 233 00:13:19,240 --> 00:13:21,439 Speaker 1: do that. And then at one point they realized, like, oh, 234 00:13:21,520 --> 00:13:25,000 Speaker 1: in addition to selling the books, we can sell whatever 235 00:13:25,080 --> 00:13:28,160 Speaker 1: compute capacity we don't need ourselves. 236 00:13:29,040 --> 00:13:32,000 Speaker 4: In this version of the story, Amazon Web Services kind 237 00:13:32,000 --> 00:13:34,880 Speaker 4: of started as a side hustle with Amazon renting out 238 00:13:34,880 --> 00:13:37,880 Speaker 4: the extra computer capacity that they had to other companies. 239 00:13:38,440 --> 00:13:42,400 Speaker 4: Fortune magazine did an interview with Amazon's current CEO, Andy Jasse, 240 00:13:42,679 --> 00:13:45,520 Speaker 4: who started out as the head of Amazon Web Services. 241 00:13:45,920 --> 00:13:48,480 Speaker 4: He said that in the early two thousands, Amazon was 242 00:13:48,480 --> 00:13:50,480 Speaker 4: starting to run in the walls when they were developing 243 00:13:50,480 --> 00:13:53,160 Speaker 4: features for the site. Remember at this point they were 244 00:13:53,200 --> 00:13:56,640 Speaker 4: still mainly thought of as a bookseller, so they started 245 00:13:56,640 --> 00:13:59,560 Speaker 4: building out internal features for their own programmers to use, 246 00:14:00,040 --> 00:14:02,520 Speaker 4: and they realized that they could sell that infrastructure to 247 00:14:02,600 --> 00:14:03,320 Speaker 4: other companies. 248 00:14:04,480 --> 00:14:06,560 Speaker 1: So that is like the long short of it. There's 249 00:14:06,600 --> 00:14:08,560 Speaker 1: a lot more nuanced there, and I think some really 250 00:14:08,600 --> 00:14:11,000 Speaker 1: great Internet historians are going to be picking through that 251 00:14:11,320 --> 00:14:14,160 Speaker 1: to bring the inside story. What is definitely true is 252 00:14:14,200 --> 00:14:15,160 Speaker 1: that they were the first. 253 00:14:15,440 --> 00:14:18,480 Speaker 4: In two thousand and six, Amazon Web Services was born, 254 00:14:18,640 --> 00:14:21,880 Speaker 4: and it was instantly popular. They were solving for a 255 00:14:21,920 --> 00:14:25,440 Speaker 4: problem that companies were already starting to have before this. 256 00:14:25,720 --> 00:14:29,000 Speaker 4: Starting an app company could be really risky, partially because 257 00:14:29,000 --> 00:14:32,000 Speaker 4: of how unpredictable the traffic demands could be. Let's say 258 00:14:32,040 --> 00:14:34,360 Speaker 4: you start up a company and your online service just 259 00:14:34,440 --> 00:14:37,960 Speaker 4: goes viral overnight. Well, if you didn't buy enough service 260 00:14:38,000 --> 00:14:39,960 Speaker 4: ahead of time, you can't keep up with demand and 261 00:14:39,960 --> 00:14:42,560 Speaker 4: you're screwed. But if you did buy a bunch of 262 00:14:42,600 --> 00:14:45,560 Speaker 4: servers and your company isn't doing so well, well, now 263 00:14:45,600 --> 00:14:47,760 Speaker 4: you're stuck with a building full of servers and you're 264 00:14:47,760 --> 00:14:51,640 Speaker 4: wasting money, and now you're really screwed. What AWS did 265 00:14:51,800 --> 00:14:54,920 Speaker 4: was let companies rent that computing space and power as 266 00:14:54,920 --> 00:14:58,320 Speaker 4: opposed to having to buy it outright themselves. Basically, they 267 00:14:58,440 --> 00:15:00,880 Speaker 4: let them do what a lot of us do, pay 268 00:15:00,920 --> 00:15:04,400 Speaker 4: Amazon a subscription fee. If your business gets popular, just 269 00:15:04,520 --> 00:15:06,840 Speaker 4: rent more space. If you have a slow month or two, 270 00:15:07,120 --> 00:15:09,440 Speaker 4: you can scale it back. And on top of that, 271 00:15:09,800 --> 00:15:12,840 Speaker 4: AWS had software development tools, which meant that you could 272 00:15:12,840 --> 00:15:15,040 Speaker 4: build on top of those tools rather than having to 273 00:15:15,080 --> 00:15:18,440 Speaker 4: start to scratch. This was huge for the web two 274 00:15:18,440 --> 00:15:21,360 Speaker 4: point zero boom. It allowed for a lot of experimentation 275 00:15:21,680 --> 00:15:24,920 Speaker 4: without as much financial risk, which in some ways felt 276 00:15:24,920 --> 00:15:27,520 Speaker 4: like a good thing, but it also meant that the 277 00:15:27,560 --> 00:15:30,680 Speaker 4: Internet became really reliant on Amazon. 278 00:15:32,480 --> 00:15:35,720 Speaker 1: Companies have gone from largely running their own servers like 279 00:15:35,760 --> 00:15:39,360 Speaker 1: physical computers in their own garages, so to speak, to 280 00:15:39,440 --> 00:15:43,720 Speaker 1: relying on the servers of AWS. But maybe even more importantly, 281 00:15:43,840 --> 00:15:48,040 Speaker 1: like we've also gone from individual companies building software from 282 00:15:48,120 --> 00:15:53,680 Speaker 1: scratch to compiling it from various building blocks provided by 283 00:15:53,840 --> 00:15:55,640 Speaker 1: other software companies. 284 00:15:56,080 --> 00:15:57,479 Speaker 3: What you then get. 285 00:15:57,440 --> 00:16:02,080 Speaker 1: Is that you create this really complex web of dependencies 286 00:16:02,120 --> 00:16:06,200 Speaker 1: and interconnections where like every piece of software relies on 287 00:16:06,360 --> 00:16:09,280 Speaker 1: another piece of software which relies on another piece of software. 288 00:16:10,320 --> 00:16:13,800 Speaker 4: However it might have started, AWS is not a side hustle. 289 00:16:13,840 --> 00:16:16,880 Speaker 4: Now it brought in almost forty billion dollars in operating 290 00:16:16,920 --> 00:16:20,200 Speaker 4: income last year. That's over half of what Amazon made. 291 00:16:20,480 --> 00:16:23,080 Speaker 4: Amazon beat Google and Microsoft to the punch here, but 292 00:16:23,240 --> 00:16:26,920 Speaker 4: both those companies eventually offered their own competing services, and 293 00:16:27,000 --> 00:16:30,440 Speaker 4: now we have this environment of three companies controlling the 294 00:16:30,440 --> 00:16:33,760 Speaker 4: cloud that houses the Internet, and what was first just 295 00:16:33,840 --> 00:16:37,240 Speaker 4: an extension of basically using someone else's computer for storage 296 00:16:37,440 --> 00:16:40,360 Speaker 4: has turned into something much much bigger. 297 00:16:42,880 --> 00:16:45,000 Speaker 1: I think that is the next sort of step that 298 00:16:45,040 --> 00:16:48,000 Speaker 1: we need to take, where it's like, in what situation 299 00:16:48,600 --> 00:16:53,640 Speaker 1: would you accept, you know, only three companies being responsible 300 00:16:53,960 --> 00:16:59,360 Speaker 1: for all of agriculture. We would immediately see how big 301 00:16:59,400 --> 00:17:03,080 Speaker 1: of a choke hold that would be. Because we understand 302 00:17:03,120 --> 00:17:06,720 Speaker 1: how important food is for our daily sustenance, we need 303 00:17:06,760 --> 00:17:10,640 Speaker 1: to start changing our perception of what the digital is, 304 00:17:11,119 --> 00:17:12,480 Speaker 1: because it is everything. 305 00:17:13,280 --> 00:17:17,560 Speaker 4: The food comparison is really interesting, Like you want a tomato, 306 00:17:18,240 --> 00:17:21,600 Speaker 4: you want a hamburger, you want, I don't know, man, 307 00:17:21,720 --> 00:17:24,040 Speaker 4: some spaghetti, all of those. I don't know why I 308 00:17:24,040 --> 00:17:26,040 Speaker 4: thought of those three that maybe I'm hungry right now. 309 00:17:26,320 --> 00:17:34,000 Speaker 4: But all of those three things are you know, if 310 00:17:34,240 --> 00:17:36,719 Speaker 4: one company goes down, if something is one of those 311 00:17:36,760 --> 00:17:39,399 Speaker 4: companies has a problem, if one of those companies is 312 00:17:39,440 --> 00:17:43,199 Speaker 4: not safe for some reason, starts slacking on safety regulations 313 00:17:43,240 --> 00:17:46,320 Speaker 4: or whatever, you just you don't get one of those 314 00:17:46,320 --> 00:17:48,080 Speaker 4: three or two of those three, or all of those 315 00:17:48,119 --> 00:17:50,480 Speaker 4: three or whatever. You don't eat, man, Yeah, you just 316 00:17:50,520 --> 00:17:51,280 Speaker 4: don't eat. 317 00:17:52,040 --> 00:17:54,800 Speaker 1: There is a reason why these three keep getting bigger 318 00:17:54,840 --> 00:17:57,159 Speaker 1: because they are also the only ones with like that 319 00:17:57,280 --> 00:18:01,800 Speaker 1: amount of money to put into something that loses its 320 00:18:01,880 --> 00:18:06,399 Speaker 1: value quite quickly, right like servers infrastructure. It's expensive to build, 321 00:18:06,840 --> 00:18:09,959 Speaker 1: it's expensive to maintain, and you have to switch it 322 00:18:10,280 --> 00:18:14,760 Speaker 1: out every five to seven years. I've spoken to vcs 323 00:18:14,920 --> 00:18:17,600 Speaker 1: for research purposes just get a sense of you know, 324 00:18:17,960 --> 00:18:21,080 Speaker 1: when you get a startup and you look at their proposition, 325 00:18:21,240 --> 00:18:23,879 Speaker 1: and for instance, they would say like, oh, actually, we 326 00:18:23,960 --> 00:18:26,840 Speaker 1: either want to build our own set of servers, our 327 00:18:26,880 --> 00:18:30,000 Speaker 1: own mini cloud, or we would want to work with 328 00:18:30,080 --> 00:18:34,399 Speaker 1: multiple partners. And these vcs just like burst out laughing 329 00:18:34,480 --> 00:18:36,840 Speaker 1: at just the proposition of that, because they were like, well, 330 00:18:37,000 --> 00:18:40,520 Speaker 1: if I'm putting my money into this and they're going 331 00:18:40,600 --> 00:18:43,959 Speaker 1: to spend it on like something that gives me an 332 00:18:43,960 --> 00:18:47,919 Speaker 1: overturn when they could just do it with AWS or 333 00:18:47,960 --> 00:18:50,720 Speaker 1: whomever's going to do it cheaper and better, at least 334 00:18:50,720 --> 00:18:53,480 Speaker 1: from their perspective, I was just never fund I would 335 00:18:53,520 --> 00:18:56,200 Speaker 1: never fund something that would also require my money to 336 00:18:56,280 --> 00:18:58,440 Speaker 1: put into machines. 337 00:18:57,960 --> 00:18:59,480 Speaker 3: That are going to die in the next five years. 338 00:18:59,520 --> 00:19:01,840 Speaker 3: It's a bad investment right now. 339 00:19:01,960 --> 00:19:05,560 Speaker 4: Smaller companies really can't compete with these giants. So you 340 00:19:05,640 --> 00:19:07,880 Speaker 4: could argue here that it's better that we have these 341 00:19:07,920 --> 00:19:11,280 Speaker 4: big tech companies monopolizing this. They have more resources, and 342 00:19:11,320 --> 00:19:13,520 Speaker 4: they can afford to maintain the data centers and keep 343 00:19:13,520 --> 00:19:16,000 Speaker 4: them working, and maybe that is a way to keep 344 00:19:16,040 --> 00:19:20,240 Speaker 4: things cheap and reliable. But for Corent, that's not a 345 00:19:20,280 --> 00:19:21,159 Speaker 4: good enough reason. 346 00:19:23,359 --> 00:19:25,760 Speaker 1: The thing that I think is important to remember is 347 00:19:25,800 --> 00:19:32,479 Speaker 1: that these technical outages have political consequences. So it is 348 00:19:32,640 --> 00:19:37,920 Speaker 1: really important to politicize what is happening here, because it's 349 00:19:38,040 --> 00:19:39,840 Speaker 1: very easy for people to say like, oh, it's just 350 00:19:39,840 --> 00:19:42,040 Speaker 1: a technical mistake, and we should be allowed to make 351 00:19:42,080 --> 00:19:46,040 Speaker 1: technical mistakes. But that is not an argument that we 352 00:19:46,080 --> 00:19:50,400 Speaker 1: would accept from, you know, the company that operates our 353 00:19:50,480 --> 00:19:54,439 Speaker 1: water or our electricity, for them to just be like, 354 00:19:54,480 --> 00:19:56,840 Speaker 1: I'm sorry, it's winter, but for the next twelve hours 355 00:19:57,320 --> 00:20:00,199 Speaker 1: no heat in your house or no clean water, so 356 00:20:00,600 --> 00:20:04,240 Speaker 1: your kids are thirsty. Too bad. And somehow with these companies, 357 00:20:04,280 --> 00:20:07,840 Speaker 1: we've decided that they get a pass or like maybe 358 00:20:07,880 --> 00:20:11,399 Speaker 1: even better, like they've just lobbied such that there isn't 359 00:20:11,480 --> 00:20:14,760 Speaker 1: that much or enough at least from my perspective oversight 360 00:20:15,000 --> 00:20:19,560 Speaker 1: on you know, them making these mistakes and that having 361 00:20:19,720 --> 00:20:22,800 Speaker 1: really potentially life altering consequences. 362 00:20:22,359 --> 00:20:25,080 Speaker 3: And that I very much see as a democratic deficit. 363 00:20:26,680 --> 00:20:29,000 Speaker 4: So what can we do about this? How do we 364 00:20:29,040 --> 00:20:31,640 Speaker 4: take back computing power from the hands of a few 365 00:20:31,680 --> 00:20:44,479 Speaker 4: big corporations. That's after the break The same week as 366 00:20:44,520 --> 00:20:47,920 Speaker 4: the AWS outage last month, political leaders in Europe met 367 00:20:47,920 --> 00:20:50,360 Speaker 4: in Brussels for a summit, and one of the items 368 00:20:50,359 --> 00:20:53,720 Speaker 4: on the agenda was this idea of European digital sovereignty, 369 00:20:54,080 --> 00:20:57,040 Speaker 4: basically figuring out a way to untangle Europe from being 370 00:20:57,119 --> 00:20:59,600 Speaker 4: so reliant on these big US tech companies. 371 00:21:00,160 --> 00:21:03,199 Speaker 1: I'm actually quite critical of people who are very focused 372 00:21:03,240 --> 00:21:06,320 Speaker 1: on the European digital sovereignty because I'm like, well, you know, 373 00:21:06,359 --> 00:21:08,720 Speaker 1: the world's a big place, and you just you know 374 00:21:08,800 --> 00:21:11,680 Speaker 1: what ford Europe is where you put your digital borders, 375 00:21:11,680 --> 00:21:15,000 Speaker 1: like that's not good enough. The perspective needs to be. 376 00:21:14,960 --> 00:21:16,879 Speaker 3: One of global solidarity always. 377 00:21:17,080 --> 00:21:22,880 Speaker 1: And I have been in the room with politicians representing 378 00:21:23,000 --> 00:21:27,080 Speaker 1: you know, the Greens or left wing parties back home, 379 00:21:27,720 --> 00:21:32,359 Speaker 1: who without irony, have said, oh, well, you know, we 380 00:21:32,480 --> 00:21:35,679 Speaker 1: need to start looking for alternatives. But the only ones 381 00:21:35,720 --> 00:21:38,440 Speaker 1: that are perhaps available to us, or that have experience 382 00:21:38,520 --> 00:21:40,639 Speaker 1: running these kind of networks and that kind of skill 383 00:21:41,359 --> 00:21:45,800 Speaker 1: are defense contractors. And I would be sitting there picking 384 00:21:45,840 --> 00:21:49,760 Speaker 1: my job from the floor, being like, wait, the choice 385 00:21:49,800 --> 00:21:56,760 Speaker 1: that we are giving ourselves here is AWS or defense contractors, Right, 386 00:21:56,840 --> 00:21:59,879 Speaker 1: that is a rock and a hard place. We have 387 00:22:00,080 --> 00:22:01,200 Speaker 1: to think of something else that. 388 00:22:01,160 --> 00:22:04,240 Speaker 4: Is not this Instead of looking to Silicon Valley or 389 00:22:04,240 --> 00:22:06,879 Speaker 4: the military for the answer, Corent says, we should be 390 00:22:06,920 --> 00:22:09,920 Speaker 4: thinking about nonprofits and universities for ideas here. 391 00:22:10,359 --> 00:22:14,720 Speaker 1: Now, it's not impossible. Wikipedia does it. They specifically have 392 00:22:14,800 --> 00:22:18,000 Speaker 1: a setup that has them. I want to say, like 393 00:22:18,080 --> 00:22:23,040 Speaker 1: across seven different data centers across numerous continents, and none 394 00:22:23,080 --> 00:22:25,040 Speaker 1: of them are the Big Three. But that would be 395 00:22:25,160 --> 00:22:27,760 Speaker 1: much harder to do for something like signal right, where 396 00:22:28,200 --> 00:22:32,560 Speaker 1: latency and like constant uptime is much more visible when 397 00:22:32,600 --> 00:22:33,200 Speaker 1: it doesn't work. 398 00:22:33,240 --> 00:22:35,280 Speaker 3: Then with something like Wikipedia. 399 00:22:35,400 --> 00:22:40,000 Speaker 1: There are alternatives, but it requires us to also change 400 00:22:40,000 --> 00:22:43,879 Speaker 1: our expectations, and it requires the people more importantly, like 401 00:22:43,920 --> 00:22:46,200 Speaker 1: who put money into this, who will literally have skin 402 00:22:46,280 --> 00:22:49,879 Speaker 1: in the game, to change theirs. So it's cultural change 403 00:22:49,920 --> 00:22:53,800 Speaker 1: as well. It's not only technical. Locally, we work with 404 00:22:53,920 --> 00:22:56,800 Speaker 1: small medium enterprises to see, like, you know, what they 405 00:22:56,840 --> 00:23:00,280 Speaker 1: are thinking of the universities in the Netherlands of taking 406 00:23:00,400 --> 00:23:04,560 Speaker 1: some really interesting steps where they have all come together 407 00:23:05,000 --> 00:23:08,000 Speaker 1: with like local providers to say like, okay, well we 408 00:23:08,040 --> 00:23:11,480 Speaker 1: are roughly the same kind of institute. Here are the 409 00:23:11,480 --> 00:23:14,879 Speaker 1: needs that we all have. Could you build us something 410 00:23:14,920 --> 00:23:17,040 Speaker 1: that is like fit for purpose for universities. 411 00:23:19,200 --> 00:23:21,560 Speaker 4: So there is a kind of precedent for what Kooran's 412 00:23:21,560 --> 00:23:24,639 Speaker 4: advocating here. And the day we did this interview was 413 00:23:24,680 --> 00:23:26,120 Speaker 4: also an important anniversary. 414 00:23:26,920 --> 00:23:28,200 Speaker 3: Today is the. 415 00:23:28,160 --> 00:23:32,720 Speaker 1: Twenty nine of October, which is Arpennet Launch Day. Arponnet 416 00:23:32,760 --> 00:23:36,800 Speaker 1: is the precursor to the Internet, and in October twenty ninth, 417 00:23:36,920 --> 00:23:41,360 Speaker 1: sixty nine, the first message was sent over the arpennet. 418 00:23:41,680 --> 00:23:45,040 Speaker 4: So the Arpaennet was basically a pre Internet network that 419 00:23:45,119 --> 00:23:48,120 Speaker 4: was developed as a collaboration between universities and the US 420 00:23:48,200 --> 00:23:50,480 Speaker 4: Department of Defense. Way back in the sixties. 421 00:23:50,960 --> 00:23:54,639 Speaker 1: And it's interesting because like that version of the Internet 422 00:23:54,960 --> 00:23:58,600 Speaker 1: was designed to be resilient. It was designed to write 423 00:23:58,600 --> 00:24:01,320 Speaker 1: around problems or failure like the ones that we saw 424 00:24:01,400 --> 00:24:05,080 Speaker 1: with WS. It was designed to fundamentally like not have 425 00:24:05,240 --> 00:24:08,040 Speaker 1: choke points and concentration points. 426 00:24:08,520 --> 00:24:12,360 Speaker 4: And so you're you're taking us back to ourpenet. 427 00:24:14,480 --> 00:24:16,560 Speaker 1: I can't believe that that is where we Yeah, yeah, 428 00:24:16,600 --> 00:24:17,639 Speaker 1: it's full circle. 429 00:24:17,800 --> 00:24:19,600 Speaker 4: No, but I mean, isn't that what's happening is I mean, 430 00:24:19,880 --> 00:24:24,639 Speaker 4: we're talking about universities without a profit motive, which is 431 00:24:24,760 --> 00:24:27,399 Speaker 4: really important here, without a profit motive, figuring out ways 432 00:24:27,400 --> 00:24:32,760 Speaker 4: to connect to each other with no singular choke point, 433 00:24:33,320 --> 00:24:34,919 Speaker 4: and then figuring out, okay, we I think we've got 434 00:24:34,920 --> 00:24:38,760 Speaker 4: assistant that works, everybody else copy it, but that is. 435 00:24:38,680 --> 00:24:42,640 Speaker 1: Still operating from the assumption that we need to. 436 00:24:42,560 --> 00:24:44,400 Speaker 3: Be moving to a world of more compute. 437 00:24:45,080 --> 00:24:51,560 Speaker 1: And some of the people that I really take inspiration 438 00:24:51,720 --> 00:24:56,600 Speaker 1: from in the field are you know, in academics, researchers, colleagues, 439 00:24:56,920 --> 00:25:02,080 Speaker 1: but also like artists and organizers who are building technologies 440 00:25:02,119 --> 00:25:07,200 Speaker 1: but also ideas and also political worlds that provide actual alternatives. 441 00:25:07,480 --> 00:25:10,120 Speaker 1: And two sort of examples that that I think might 442 00:25:10,160 --> 00:25:13,240 Speaker 1: be interesting is the Critical Infrastructure Lab at the University 443 00:25:13,280 --> 00:25:17,399 Speaker 1: of Amsterdam and the Institute for Technology in the Public Interest, 444 00:25:17,440 --> 00:25:19,840 Speaker 1: which is based out of Brussels, and they are doing 445 00:25:20,040 --> 00:25:24,680 Speaker 1: super fascinating stuff. They are literally like building compostable data centers. 446 00:25:25,040 --> 00:25:28,280 Speaker 1: They are generating energy out of mud just to show 447 00:25:28,520 --> 00:25:30,680 Speaker 1: how hard it is to get energy and like how 448 00:25:30,760 --> 00:25:34,399 Speaker 1: much of it data centers consume. They are creating like 449 00:25:34,520 --> 00:25:38,080 Speaker 1: cloud abolitionists agendas, and that is the kind of thinking 450 00:25:38,280 --> 00:25:41,320 Speaker 1: and action that we also need. We need people will 451 00:25:41,359 --> 00:25:45,040 Speaker 1: actually question the premises and not just try and improve 452 00:25:45,560 --> 00:25:46,560 Speaker 1: what we have already. 453 00:25:47,440 --> 00:25:49,439 Speaker 4: I don't know if this is recency bias, but we 454 00:25:49,480 --> 00:25:51,520 Speaker 4: set up this interview I wanted to speak to you 455 00:25:51,640 --> 00:25:57,080 Speaker 4: about AWS as we're talking right now, Microsoft's Azure is down. 456 00:25:58,119 --> 00:26:00,520 Speaker 4: Is this just recency bias or is this stuff happening 457 00:26:00,520 --> 00:26:02,600 Speaker 4: more frequently and are we going to be seeing this 458 00:26:02,680 --> 00:26:03,680 Speaker 4: happen more in the future. 459 00:26:04,320 --> 00:26:07,200 Speaker 1: Yes, yes, and yes, I mean we do tend to 460 00:26:07,240 --> 00:26:09,600 Speaker 1: see the things more often when we're focused on them, 461 00:26:10,040 --> 00:26:14,439 Speaker 1: and this thing happens more often than people would like 462 00:26:14,480 --> 00:26:17,920 Speaker 1: to admit. The skill of the impact tends to generate 463 00:26:18,080 --> 00:26:21,639 Speaker 1: more or less attention depending on who it specifically hits 464 00:26:21,840 --> 00:26:24,560 Speaker 1: and yes, are we likely to see this more in 465 00:26:24,600 --> 00:26:29,280 Speaker 1: the future. Absolutely, Will the impact be bigger every time? Yes, 466 00:26:29,560 --> 00:26:31,800 Speaker 1: because one of the biggest things that we have on 467 00:26:31,840 --> 00:26:34,840 Speaker 1: our radar is that there are a lot of sectors 468 00:26:34,840 --> 00:26:38,600 Speaker 1: that are not yet digitized who are immediately going to 469 00:26:38,640 --> 00:26:42,879 Speaker 1: make the jump to the cloud. So whatever the next 470 00:26:42,920 --> 00:26:47,119 Speaker 1: iteration of this is, just because of that, the impact 471 00:26:47,160 --> 00:26:49,200 Speaker 1: will likely be even bigger than the last one. 472 00:26:49,680 --> 00:26:51,800 Speaker 4: Yeah, I mean, I'm just thinking about how excited I 473 00:26:52,040 --> 00:26:55,800 Speaker 4: was when finally the trains and buses here in LA, 474 00:26:56,400 --> 00:26:58,719 Speaker 4: I was able to not have to carry around that 475 00:26:58,880 --> 00:27:02,359 Speaker 4: dumb card and I could use my app. And I 476 00:27:02,359 --> 00:27:04,399 Speaker 4: mean this is a real thing because also here in 477 00:27:04,520 --> 00:27:08,240 Speaker 4: LA they are introducing police at the entrance points. I 478 00:27:08,240 --> 00:27:10,080 Speaker 4: would very much like to not get harassed by the 479 00:27:10,119 --> 00:27:12,280 Speaker 4: police because all of a sudden my phone isn't working 480 00:27:12,760 --> 00:27:15,399 Speaker 4: to unlock my app so that I can get on 481 00:27:15,440 --> 00:27:17,280 Speaker 4: the train, and they think that I'm lying to them 482 00:27:17,359 --> 00:27:18,960 Speaker 4: that I have paid for my fare. 483 00:27:19,520 --> 00:27:19,760 Speaker 2: Yep. 484 00:27:20,520 --> 00:27:22,760 Speaker 4: It isn't just that these companies are all at risk 485 00:27:22,800 --> 00:27:26,000 Speaker 4: of experiencing another outage. A lot of Korn's work is 486 00:27:26,040 --> 00:27:28,600 Speaker 4: looking at all the ways that these companies control how 487 00:27:28,640 --> 00:27:31,240 Speaker 4: we live, our lives and what it means to have 488 00:27:31,359 --> 00:27:35,080 Speaker 4: so much power. I want to point out something that 489 00:27:35,200 --> 00:27:39,160 Speaker 4: AWS did that I think a lot of people maybe 490 00:27:39,200 --> 00:27:43,520 Speaker 4: we're a fan of. Actually, in twenty twenty one, AWS 491 00:27:43,840 --> 00:27:48,240 Speaker 4: pulled Parlor from being able to use their starvers. Just 492 00:27:48,280 --> 00:27:51,760 Speaker 4: for context here, Parlor think of it as it's another 493 00:27:51,840 --> 00:27:56,960 Speaker 4: social media app which was pretty explicitly right wing. Let's 494 00:27:57,000 --> 00:27:59,480 Speaker 4: just say that a lot of far right personalities have 495 00:27:59,560 --> 00:28:05,680 Speaker 4: pretty big followings there. And after January sixth, a little 496 00:28:05,680 --> 00:28:09,600 Speaker 4: bit after January sixth, AWS said, hey, listen, we're not 497 00:28:09,640 --> 00:28:13,399 Speaker 4: hosting this anymore. And I think they told Parlor that 498 00:28:13,440 --> 00:28:17,320 Speaker 4: have found ninety eight posts on the site that encouraged violence. 499 00:28:17,800 --> 00:28:21,200 Speaker 4: A lot of people said, good Amazon, doing the right thing. 500 00:28:22,600 --> 00:28:25,600 Speaker 4: Could that potentially cut both ways? 501 00:28:26,080 --> 00:28:32,280 Speaker 1: Absolutely? At article nineteen. We have spent a decade documenting, researching, 502 00:28:32,400 --> 00:28:37,720 Speaker 1: and intervening in both corporate and government abuse of these 503 00:28:37,760 --> 00:28:42,600 Speaker 1: infrastructures to quell the scent, to make it harder to organize, 504 00:28:42,640 --> 00:28:45,080 Speaker 1: to make it harder for human rights defenders to speak up. 505 00:28:45,280 --> 00:28:48,800 Speaker 1: What this has shown us is the specific power dynamic 506 00:28:48,960 --> 00:28:52,200 Speaker 1: of how these cloud companies get to decide who gets 507 00:28:52,200 --> 00:28:55,560 Speaker 1: to use their services and by extension, who gets to 508 00:28:55,600 --> 00:28:59,760 Speaker 1: exercise their right to online? When so few companies control 509 00:28:59,800 --> 00:29:03,960 Speaker 1: the infrastructure, they also effectively control the flow of information, 510 00:29:04,080 --> 00:29:08,480 Speaker 1: and that has view implications for our ability to organize, 511 00:29:08,560 --> 00:29:12,360 Speaker 1: to find community, to access information, to exercise your freedom 512 00:29:12,360 --> 00:29:12,920 Speaker 1: of expression. 513 00:29:13,320 --> 00:29:17,920 Speaker 4: Yeah, I mean so many people rely on signal signal, 514 00:29:18,000 --> 00:29:23,240 Speaker 4: relies on aws. Yeah, and what happens if for some 515 00:29:23,280 --> 00:29:29,640 Speaker 4: political reason, Amazon decides no more signal? Right now? I 516 00:29:29,640 --> 00:29:33,440 Speaker 4: think it's very reasonable to assume, at least at the stage, 517 00:29:33,480 --> 00:29:35,600 Speaker 4: they never do that. They never do that, Okay, what 518 00:29:35,640 --> 00:29:36,720 Speaker 4: if they do? Right? 519 00:29:36,960 --> 00:29:39,280 Speaker 1: We need to just be prepared for these kind of 520 00:29:39,320 --> 00:29:42,480 Speaker 1: scenarios and make sure that these companies have a sense 521 00:29:42,520 --> 00:29:46,040 Speaker 1: of accountability for their role. It's the Spider Man quote, 522 00:29:46,080 --> 00:29:48,240 Speaker 1: it's a great power, great responsibility. 523 00:29:48,680 --> 00:29:50,880 Speaker 4: So how do we hold that great responsibility? How do 524 00:29:50,880 --> 00:29:52,480 Speaker 4: we keep them accountable. 525 00:29:52,360 --> 00:29:54,040 Speaker 1: As individual persons? 526 00:29:53,640 --> 00:29:54,880 Speaker 3: It's very hard. 527 00:29:55,640 --> 00:29:57,360 Speaker 1: Like I know that people said like, oh, you can 528 00:29:57,480 --> 00:30:00,400 Speaker 1: start self hosting all of your stuff, and like that's possible, 529 00:30:00,440 --> 00:30:02,400 Speaker 1: But then you know, try and self host your own 530 00:30:02,440 --> 00:30:04,920 Speaker 1: email right now and then get through the Gmail filters 531 00:30:04,960 --> 00:30:07,640 Speaker 1: just forget about it. But I do think there are 532 00:30:07,760 --> 00:30:10,280 Speaker 1: other stuffs that we can and should be taking Like 533 00:30:10,520 --> 00:30:12,360 Speaker 1: I want to give a shout out to my colleague, 534 00:30:12,440 --> 00:30:15,840 Speaker 1: doctor Isa Stasi, who's a lawyer at Article nineteen who 535 00:30:15,920 --> 00:30:20,720 Speaker 1: is currently taking Microsoft to court for unfair licensing practices. Right, 536 00:30:20,840 --> 00:30:23,800 Speaker 1: there are things that people can do. It's just it's 537 00:30:23,960 --> 00:30:27,200 Speaker 1: very hard, and you know, the entirety of the weight 538 00:30:27,400 --> 00:30:30,440 Speaker 1: of that corporate power comes sparing down on you. So 539 00:30:30,640 --> 00:30:32,560 Speaker 1: you need to have the kind of you know, backbone 540 00:30:32,600 --> 00:30:33,800 Speaker 1: to take on that backbone. 541 00:30:35,800 --> 00:30:38,360 Speaker 4: So unless you're the head of some massive tech company 542 00:30:38,360 --> 00:30:42,520 Speaker 4: that's also using AWS or Microsoft Azure, Google Cloud, the 543 00:30:42,560 --> 00:30:46,240 Speaker 4: decision on de monopolizing the Internet probably isn't directly in 544 00:30:46,280 --> 00:30:49,040 Speaker 4: your hands. This is a big problem and it's not 545 00:30:49,200 --> 00:30:51,720 Speaker 4: something that any one of us as an individual is 546 00:30:51,720 --> 00:30:55,000 Speaker 4: going to fix. But at the very least, this is 547 00:30:55,120 --> 00:30:58,200 Speaker 4: another instance where we should be able to have some transparency. 548 00:30:58,640 --> 00:31:01,880 Speaker 4: Companies don't have to answer all of our questions, but 549 00:31:02,000 --> 00:31:04,880 Speaker 4: governments are a little different. We should be able to 550 00:31:04,920 --> 00:31:07,880 Speaker 4: ask our local officials what the hospital is depending on 551 00:31:07,960 --> 00:31:10,840 Speaker 4: to keep people alive, and what the backup plan is 552 00:31:11,000 --> 00:31:15,120 Speaker 4: when not if, but when that service goes offline next. 553 00:31:15,680 --> 00:31:19,000 Speaker 4: It's a small thing, but maybe that's one step toward 554 00:31:19,080 --> 00:31:24,360 Speaker 4: thinking about this stuff a little differently. Thank you so 555 00:31:24,480 --> 00:31:27,239 Speaker 4: much for listening to another episode of kill Switch. If 556 00:31:27,280 --> 00:31:28,920 Speaker 4: you want to talk to us, you could email us 557 00:31:29,000 --> 00:31:32,840 Speaker 4: at kill Switch at Kaleidoscope dot NYC, or we're also 558 00:31:32,880 --> 00:31:36,760 Speaker 4: on Instagram at kill Switch pod. And while you got 559 00:31:36,760 --> 00:31:38,760 Speaker 4: your phone out, you know, maybe leave us a review. 560 00:31:39,160 --> 00:31:41,720 Speaker 4: It helps other people find the show, which helps us 561 00:31:41,880 --> 00:31:44,800 Speaker 4: keep doing our thing. Kill Switch is hosted by Me 562 00:31:45,080 --> 00:31:48,880 Speaker 4: Dexter Thomas. It's produced by Shena Ozaki, Dar Luck Potts, 563 00:31:49,080 --> 00:31:51,960 Speaker 4: and Julia Nutter. Our theme song is by me and 564 00:31:52,040 --> 00:31:56,040 Speaker 4: Kyle Murdoch, and Kyle also mixes the show Kaleidoscope. Our 565 00:31:56,080 --> 00:32:00,560 Speaker 4: executive producers are Osra Lashin, Mangesh Hajikadur and k Osbourne 566 00:32:01,000 --> 00:32:04,760 Speaker 4: from iHeart. Our executive producers our Katrina Norville and Nikki 567 00:32:04,800 --> 00:32:08,320 Speaker 4: etoor Oh. And by the way, I asked Karin if 568 00:32:08,320 --> 00:32:10,800 Speaker 4: she had anything she wanted to add, and she said 569 00:32:10,840 --> 00:32:11,200 Speaker 4: she did. 570 00:32:12,280 --> 00:32:13,960 Speaker 1: People in the past have asked me, like, you know, 571 00:32:14,040 --> 00:32:17,440 Speaker 1: what kind of power or force does it take to 572 00:32:17,480 --> 00:32:21,920 Speaker 1: push back on the sort of self propelling force of 573 00:32:22,000 --> 00:32:25,800 Speaker 1: these data centers? And I do think that, like coming 574 00:32:25,840 --> 00:32:30,480 Speaker 1: to the realization that we cannot keep growing our computing 575 00:32:30,520 --> 00:32:34,520 Speaker 1: industry like this without us becoming the victims of it 576 00:32:34,560 --> 00:32:37,680 Speaker 1: is really important. One of the big things that we 577 00:32:37,760 --> 00:32:41,240 Speaker 1: know about data centers is that it takes huge amounts 578 00:32:41,280 --> 00:32:46,200 Speaker 1: of energy, huge amounts of water, land, et cetera. So like, 579 00:32:46,400 --> 00:32:49,200 Speaker 1: even for the most sort of hard nos, like I'm 580 00:32:49,240 --> 00:32:52,680 Speaker 1: all about competition, and like I want to whatever make 581 00:32:52,760 --> 00:32:55,320 Speaker 1: money out of this, Like, yes, I'm sure you do. 582 00:32:55,400 --> 00:32:58,560 Speaker 1: Would you also like to be alive because those two 583 00:32:58,600 --> 00:33:01,240 Speaker 1: things in the long run not possible. 584 00:33:02,440 --> 00:33:03,880 Speaker 4: If you want to hear more on that, we do 585 00:33:03,960 --> 00:33:05,760 Speaker 4: have an episode on the time that I tried to 586 00:33:05,760 --> 00:33:09,080 Speaker 4: figure out whether or not my embarrassingly prolific use of 587 00:33:09,160 --> 00:33:13,240 Speaker 4: chat GPT is better or worse than driving a gas 588 00:33:13,280 --> 00:33:16,680 Speaker 4: guzzling car. And we'll leave that as always in the 589 00:33:16,720 --> 00:33:21,840 Speaker 4: show notes. Anyway, catch all the next one.