WEBVTT - Hard Fork's Kevin Roose and Casey Newton & Cambridge Analytica (Part I) 0:00:01.280 --> 0:00:03.800 In some very real ways. Cambridge Analytica changed my life. 0:00:03.960 --> 0:00:07.200 Casey is the only person in America who benefited from. 0:00:09.320 --> 0:00:10.320 This was great for me. 0:00:22.120 --> 0:00:25.720 Welcome to SNAFU, the podcast about history's greatest screw ups. 0:00:25.720 --> 0:00:29.760 I'm your host ed helms and each episode, as you know, 0:00:30.040 --> 0:00:32.840 I cover an enormous grew up from history to see 0:00:33.200 --> 0:00:37.360 what we can learn from humanity's biggest mistakes. My guests 0:00:37.400 --> 0:00:43.680 today are truly two of like just the most awesome people. Truly, 0:00:43.800 --> 0:00:46.640 I am so excited. We have Kevin Russ, who is 0:00:46.800 --> 0:00:49.600 an award winning tech columnist for The New York Times, 0:00:49.680 --> 0:00:53.800 frequently covering Silicon Valley, social media and tech developments. He's 0:00:53.880 --> 0:00:58.600 also the author of several books, including The AGI Chronicles, 0:00:58.680 --> 0:01:03.040 the inside story of the Race to create an Artificial Superintelligence, 0:01:03.600 --> 0:01:07.200 releasing later this year. Welcome Kevin Ruse. 0:01:07.720 --> 0:01:08.120 Thank you. 0:01:08.440 --> 0:01:11.959 And we also have not just Kevin, we also have 0:01:12.040 --> 0:01:15.360 Casey Newton, who is the founder and editor of Platformer, 0:01:15.520 --> 0:01:19.160 a publication about the intersection of democracy and tech, and 0:01:19.480 --> 0:01:23.800 was previously a senior editor at The Verge. Together these 0:01:23.800 --> 0:01:29.360 two come together in a podcasting voltron as the podcast 0:01:29.440 --> 0:01:33.039 hard Fork, which is brilliant. It's hugely popular. It is 0:01:33.080 --> 0:01:36.000 incredibly successful. It is a New York Times podcast, and 0:01:36.080 --> 0:01:40.080 it is about the rapidly changing tech world, and somehow 0:01:40.080 --> 0:01:44.800 it's it's incredibly grounded and informative even when these things 0:01:44.840 --> 0:01:48.000 feel dark and scary, and it's also insanely funny and entertaining. Welcome, 0:01:48.160 --> 0:01:49.240 Casey and Kevin. 0:01:49.640 --> 0:01:51.720 Thank you, ed, it's great to be here. Thank you 0:01:51.760 --> 0:01:54.240 for those kind words. And I'm sorry Kevin's bios so long. 0:01:54.240 --> 0:01:55.720 We're trying to cut that down a little. 0:01:57.960 --> 0:01:59.480 I just keep adding fake things to it. 0:01:59.520 --> 0:02:03.560 Olympic Gold medalists, Nobel Prize winner. 0:02:03.800 --> 0:02:05.760 That's all good. I think I think anyone can claim 0:02:05.800 --> 0:02:10.080 a Nobel at this point. What is it? I'm just 0:02:10.120 --> 0:02:13.440 so curious as a fan of hard Fork, like, and 0:02:13.480 --> 0:02:15.480 I'm sure you get asked this a lot, but were 0:02:15.480 --> 0:02:19.119 you guys friends before hard Fork or did you sort 0:02:19.160 --> 0:02:22.000 of come together? Was this sort of like a more 0:02:22.040 --> 0:02:24.880 of a business venture that and then because you do 0:02:25.080 --> 0:02:27.960 have the dynamic of people who seemed to just have 0:02:28.080 --> 0:02:30.920 known each other a long time, it's a very easy vibe. 0:02:31.480 --> 0:02:33.880 I think that the answer is that we were friends 0:02:33.919 --> 0:02:36.120 who have become much closer friends since we started doing 0:02:36.160 --> 0:02:38.200 the show. Like Kevin and I knew each other from 0:02:38.240 --> 0:02:41.040 San Francisco tech reporting circles. We would see each other 0:02:41.120 --> 0:02:43.280 all over town as we were, you know, covering this 0:02:43.320 --> 0:02:46.639 Google event, that Apple event. But both of us really 0:02:46.639 --> 0:02:49.359 wanted to start a podcast, and when we looked around, 0:02:49.639 --> 0:02:51.520 and the truth is, both of us tried very hard 0:02:51.520 --> 0:02:53.520 to find someone else to do the show with, and 0:02:53.560 --> 0:02:55.520 we just kept coming back to each other. We were 0:02:55.520 --> 0:02:57.840 just kept coming back, like, I think this is honestly 0:02:57.880 --> 0:03:00.520 the only person I could imagine doing this way. And like, 0:03:00.560 --> 0:03:03.200 almost five years later since we started having that conversation, like, 0:03:03.280 --> 0:03:04.480 I feel exactly the same way. 0:03:04.600 --> 0:03:05.520 He's the only one for me. 0:03:05.800 --> 0:03:08.720 Ah, it's a charade. We actually hate each other. This 0:03:08.760 --> 0:03:11.360 is just a sort of kabookie theater bit we do. 0:03:12.000 --> 0:03:14.000 Yeah, this is the first time we've talked outside of 0:03:14.040 --> 0:03:16.440 a hard Fork taping and sels. 0:03:18.080 --> 0:03:20.400 Oh, well, you're you're putting on a good show. I'm 0:03:20.560 --> 0:03:22.959 I'm definitely buying it. I think one of the reasons 0:03:22.960 --> 0:03:25.640 hard Fork works for me just as a fan is 0:03:25.639 --> 0:03:28.240 that I just trust you guys. I feel like you're 0:03:28.520 --> 0:03:32.680 incredibly well informed, but you also have this uncanny ability 0:03:32.720 --> 0:03:37.360 to stay calm and steady when things feel so terrifying, 0:03:37.720 --> 0:03:41.200 and I'm just wondering as a non super tech person, 0:03:41.360 --> 0:03:43.560 I just feel so confused so much. And you guys 0:03:43.600 --> 0:03:47.760 are such a steady calming force. How do you do that? 0:03:48.200 --> 0:03:50.160 How do you you're in the middle of it. It would 0:03:50.160 --> 0:03:52.000 seem like you would just be running around with your 0:03:52.000 --> 0:03:53.280 hair on fire all the time. 0:03:53.640 --> 0:03:56.440 I mean, I think both of us are optimists by nature, 0:03:57.600 --> 0:04:00.920 where both people who have observed that like, for the 0:04:00.920 --> 0:04:04.640 most part, technology has been good for people, Like we're 0:04:04.640 --> 0:04:06.520 not people who wish that we could go back to 0:04:06.600 --> 0:04:10.280 an agrarian economy and sort of do subsistence farming, like 0:04:10.560 --> 0:04:13.160 I like having an iPad, you know, And so we 0:04:13.320 --> 0:04:16.080 want to believe that as uncertain and scary as the 0:04:16.120 --> 0:04:20.040 world often is, we are going to muddle through somehow. 0:04:20.520 --> 0:04:22.200 And so, you know, while we pay a lot of 0:04:22.200 --> 0:04:24.520 attention to the many, many things that are going wrong 0:04:24.520 --> 0:04:27.479 with snawfoos, if you will, we also think there's like 0:04:27.480 --> 0:04:29.440 some pretty cool stuff happening in the world, and we 0:04:29.480 --> 0:04:31.000 like pointing that out whenever we can. 0:04:31.200 --> 0:04:36.279 I think we try to be skeptical without being cynical. Great, 0:04:36.320 --> 0:04:37.640 it would be the way I put it, you know. 0:04:37.720 --> 0:04:41.960 I think we try to be reporters and question authority 0:04:42.040 --> 0:04:46.480 and the official narratives. But we're also not sort of 0:04:46.600 --> 0:04:52.320 knee jerk reflexively cynical about the fact that some of 0:04:52.360 --> 0:04:55.919 this technology is quite good and improving in ways that 0:04:56.000 --> 0:04:59.479 should make us both excited and fearful. So that's the 0:04:59.480 --> 0:05:01.520 balance which to strike. We try to just be really honest. 0:05:01.520 --> 0:05:03.919 Some weeks we're really terrified, and you can hear that 0:05:03.960 --> 0:05:06.159 in the show. Some weeks were more excited, and hopefully 0:05:06.200 --> 0:05:06.920 you could hear that too. 0:05:07.080 --> 0:05:09.159 Well. That is very well said. And I love the 0:05:09.200 --> 0:05:13.560 distinction between skepticism and cynicism, and I feel like that's 0:05:13.640 --> 0:05:17.920 not a distinction that is made often enough, and sometimes 0:05:18.000 --> 0:05:21.479 in our own guts, like I know, I crossover. I 0:05:21.560 --> 0:05:23.680 like to think I'm just a sort of healthy skeptic, 0:05:24.000 --> 0:05:27.440 but sometimes that cynicism creeps in, and boys that get dark, 0:05:27.720 --> 0:05:30.640 oh boy. But that's why you're here, That's why hard 0:05:30.640 --> 0:05:34.440 work is there to uplift all of us, and it's 0:05:34.480 --> 0:05:37.359 why I am so thrilled to have you both on today. 0:05:37.400 --> 0:05:41.600 I've picked a doozy of a snaffoo to discuss, and 0:05:41.640 --> 0:05:45.679 it's one that you both know inside and out, which 0:05:45.720 --> 0:05:48.120 is kind of interesting. Because that's a little bit of 0:05:48.120 --> 0:05:52.680 a a flip of our typical format. Usually I'm telling 0:05:52.720 --> 0:05:56.320 the guest a snaffoo that they may or may not 0:05:56.440 --> 0:05:59.320 know anything about, and either way it's a surprise, whatever 0:05:59.360 --> 0:06:01.359 the story is. And in that way, I'm sort of 0:06:01.360 --> 0:06:05.200 the professor. They're the student. But in this case, this 0:06:05.279 --> 0:06:08.680 is a story that you both deeply reported on at 0:06:08.720 --> 0:06:12.080 the time it was happening, and you are deeply well 0:06:12.080 --> 0:06:15.040 informed about. So I am very much the student. I'm 0:06:15.080 --> 0:06:17.880 excited to I'll still be the narrative anchor. I'm going 0:06:17.960 --> 0:06:20.000 to kind of walk us through the story, but I'm 0:06:20.040 --> 0:06:23.000 so eager to learn from you, guys, and hear from 0:06:23.400 --> 0:06:27.359 you the sort of details and color that we just 0:06:27.480 --> 0:06:30.720 ordinarily wouldn't get in an episode like this. So thank 0:06:30.800 --> 0:06:33.760 you for being here, and we're about to dive into 0:06:34.640 --> 0:06:43.839 the story of Cambridge Analytica right off the bat. Guys, 0:06:44.080 --> 0:06:48.920 what is that conjure in you? Tension, excitement, humor? What 0:06:49.520 --> 0:06:50.400 is that? 0:06:50.960 --> 0:06:55.120 It makes me honestly a little bit nostalgic. Cambridge Analytica 0:06:55.240 --> 0:06:58.000 was honestly really important in my life. I had started 0:06:58.040 --> 0:07:01.320 writing a newsletter just a few months before that scandal, 0:07:01.720 --> 0:07:04.000 and I was mostly at the time writing about the 0:07:04.040 --> 0:07:06.120 backlash to Facebook that had to be gone after the 0:07:06.120 --> 0:07:09.760 twenty six US presidential election, and my newsletter was getting, 0:07:09.760 --> 0:07:11.640 you know, a little traction here, a little traction there, 0:07:11.760 --> 0:07:14.320 and then Cambridge analytic happened, and all of a sudden 0:07:14.400 --> 0:07:17.080 it felt like, to some people, the biggest story in 0:07:17.120 --> 0:07:19.760 the world. And so my newsletter grow. It wound up 0:07:19.840 --> 0:07:22.040 enabling me to like quit my job, start a podcast. 0:07:22.080 --> 0:07:24.520 So like, in the very real ways, Cambridge Analytica changed 0:07:24.520 --> 0:07:24.920 my life. 0:07:25.040 --> 0:07:27.640 Casey is the only person in America who benefited from 0:07:27.640 --> 0:07:28.800 the Cambridge. 0:07:30.280 --> 0:07:31.440 Like this was great for me. 0:07:34.040 --> 0:07:34.680 I love it. 0:07:35.160 --> 0:07:37.160 I have distant memories. 0:07:37.200 --> 0:07:39.600 I mean, I have the memory of a goldfish, so 0:07:39.640 --> 0:07:42.720 I forget things that I reported about last week. But 0:07:42.840 --> 0:07:45.600 when you say the words, I am filled with a 0:07:45.680 --> 0:07:48.800 kind of yeah, mid twenty tens nostalgia. I started thinking 0:07:48.880 --> 0:07:53.120 about the Harlem Shake, you know, and other the ice 0:07:53.160 --> 0:07:57.240 bucket challenge sort of exactly other phenomena from that era. 0:07:57.480 --> 0:07:59.800 Of course, Well, just to give the audience a little 0:07:59.800 --> 0:08:04.800 bit of grounding, the Cabridge Analytica data harvesting scandal of 0:08:04.880 --> 0:08:09.600 twenty eighteen was is considered one of the most massive 0:08:09.840 --> 0:08:14.160 and possibly worst data misuse cases in history. This was 0:08:14.240 --> 0:08:17.680 eight years ago or more. No, like what it was 0:08:17.680 --> 0:08:21.880 twenty fourteen? Just twelve years ago? What year are we in? 0:08:21.920 --> 0:08:25.679 Well, so this actually gets to a really interesting eighteen Yes, 0:08:26.000 --> 0:08:26.840 where do you start it? 0:08:27.400 --> 0:08:27.640 Yeah? 0:08:27.800 --> 0:08:27.960 So. 0:08:28.920 --> 0:08:31.560 One of the reasons why Cambridge Analytica wound up being 0:08:31.600 --> 0:08:34.920 a really weird scandal was that most of the details 0:08:35.040 --> 0:08:38.800 were known years before it turned into this conflagration, and 0:08:38.840 --> 0:08:41.440 it was only in the aftermath of Trump's election that 0:08:41.559 --> 0:08:44.520 a lot of Americans said, wait, what exactly was going on? 0:08:44.640 --> 0:08:45.679 And then it blew up. 0:08:46.160 --> 0:08:49.000 Oh boy, all right, let's dive in to kick off 0:08:49.000 --> 0:08:51.040 our snapfoo, We're going to travel back in time, like 0:08:51.080 --> 0:08:53.680 we were saying, to the mid twenty tens. So just 0:08:53.679 --> 0:08:56.080 to help reset your culture clocks. Obama was in his 0:08:56.120 --> 0:09:00.600 second term. Russia has annexed crimea. Everyone's freaking out about 0:09:00.640 --> 0:09:04.680 a bola. In the tech world, apple watches and other 0:09:04.800 --> 0:09:06.440 kinds of wearables are on the rise. 0:09:07.559 --> 0:09:07.719 Yeah. 0:09:07.880 --> 0:09:10.520 Yeah. As Kevin mentioned the ice bucket challenge, we were 0:09:10.559 --> 0:09:13.800 all boring ice water on our heads, and Instagram was 0:09:13.880 --> 0:09:17.439 primarily just photos of eggs, benedict and sunsets. The good 0:09:17.440 --> 0:09:21.800 old days. AI still felt like sci fi. Back then, 0:09:22.400 --> 0:09:26.480 chat GPT wasn't even a blip. We were young, optimistic, 0:09:27.320 --> 0:09:31.840 and really maybe a little too casual about clicking. I 0:09:32.040 --> 0:09:35.920 agree on terms and conditions that no one read. How 0:09:36.040 --> 0:09:40.719 naive we were. Was a better time now. While we 0:09:40.720 --> 0:09:43.400 were happily posting memes and baby photos, we didn't even 0:09:43.440 --> 0:09:45.439 think to worry about what was happening to all the 0:09:45.520 --> 0:09:48.480 data that we were putting out into social media or 0:09:48.679 --> 0:09:52.679 what someone could possibly do with it. But I'm kind 0:09:52.679 --> 0:09:54.600 of getting ahead of myself. Let's start with something innocent. 0:09:54.960 --> 0:09:58.960 Facebook quiz apps. Do you remember these, the harmless time 0:09:59.000 --> 0:10:03.080 wasters like what is your ice cream? Iq? Or what 0:10:03.320 --> 0:10:05.080 carb are you emotionally? 0:10:05.760 --> 0:10:07.520 Or which Harry Potter House? 0:10:08.240 --> 0:10:11.480 Exactly which Harry Potter House? Did you take any of 0:10:11.480 --> 0:10:15.400 these quizzes? Do you remember any results? I remember that 0:10:15.480 --> 0:10:17.800 I was. I was all of the carbs. Actually it 0:10:17.880 --> 0:10:21.760 turned out that I identified equally with all of them. Yeah, 0:10:21.920 --> 0:10:25.280 do we know any what was anyone in Gryffindor? 0:10:27.280 --> 0:10:29.400 I feel like heaven is definitely a slather. 0:10:31.200 --> 0:10:32.959 That's such a tough thing to say. 0:10:33.080 --> 0:10:34.200 That's a burn too. 0:10:34.520 --> 0:10:35.840 No, I love the scheme. 0:10:36.679 --> 0:10:39.560 I took one and it turns out I was. It 0:10:39.600 --> 0:10:42.000 was like which bagel are you? And I'm in Everything Bagel, 0:10:42.040 --> 0:10:44.880 which I think was supposed to be flattering, but I 0:10:44.920 --> 0:10:48.520 don't like everything Bagel. I think they're kind of gross. 0:10:50.360 --> 0:10:53.400 So these all these quizzes, they seemed kind of innocent 0:10:53.440 --> 0:10:56.040 and fun, but at least one in particular we now 0:10:56.120 --> 0:10:59.000 know was quite nefarious. And this was an app called 0:10:59.440 --> 0:11:03.440 this is Your Digital Life. Seems harmless enough, and it 0:11:03.559 --> 0:11:06.680 claimed to be for academic research. This is Your Digital 0:11:06.720 --> 0:11:09.800 Life was a personality quiz produced by a data scientist 0:11:10.160 --> 0:11:15.000 named Alexander Cogan and his company Global Science Research, which 0:11:15.080 --> 0:11:19.400 right away, I just don't trust Global Science Research. It's 0:11:20.200 --> 0:11:21.200 too much going on there. 0:11:21.360 --> 0:11:24.319 I think that's the same company from the Alien movies exactly. 0:11:24.480 --> 0:11:24.680 You know. 0:11:24.880 --> 0:11:28.520 Yeah, Paul Reiser is an executive and he's just he's 0:11:28.559 --> 0:11:32.920 a little smarmy, weird vibes. So this was in twenty fourteen, 0:11:33.480 --> 0:11:36.760 and here's where it starts getting icky. It wasn't just 0:11:36.800 --> 0:11:40.200 collecting your answers. It was also quietly scooping up data 0:11:40.280 --> 0:11:44.360 on your Facebook friends, who definitely did not sign up 0:11:44.400 --> 0:11:48.200 for that. So, thanks to a handy Facebook API loophole, 0:11:48.280 --> 0:11:53.000 the app could access information on jobs, education, location, relationship status, 0:11:53.320 --> 0:11:57.360 and liked pages. So while you were figuring out which 0:11:57.559 --> 0:12:01.640 pasta shape matched your vibe, this app was building a 0:12:01.679 --> 0:12:06.280 surprisingly detailed dossier on you, your psychology, and that of your 0:12:06.480 --> 0:12:11.920 entire digital entourage. Question, do you think that the average 0:12:11.920 --> 0:12:16.360 American at that time had any sense of how this 0:12:16.480 --> 0:12:21.320 kind of personal information that everyone was sharing just exhaustively 0:12:21.840 --> 0:12:26.000 could be turned against them or were we just high 0:12:26.080 --> 0:12:29.640 on digital narcissism? No? 0:12:29.800 --> 0:12:30.880 I mean an American. 0:12:31.120 --> 0:12:34.160 Americans are very interesting in that they will say that 0:12:34.200 --> 0:12:37.400 privacy is very important to them, but they will give 0:12:37.520 --> 0:12:39.400 up their privacy for almost anything. 0:12:39.760 --> 0:12:39.960 Right. 0:12:40.040 --> 0:12:42.560 It's like they will tell you, you know, that they 0:12:42.600 --> 0:12:44.280 don't want to, you know, put the name of their 0:12:44.360 --> 0:12:47.199 children online, but if it got them ten percent off 0:12:47.240 --> 0:12:49.520 at ross dress for less, they will absolutely give you 0:12:49.640 --> 0:12:50.400 all that and more. 0:12:50.640 --> 0:12:50.760 So. 0:12:51.080 --> 0:12:53.480 Companies are always sort of doing this weird dance where 0:12:53.480 --> 0:12:55.480 they're trying to get you to reveal as much as 0:12:55.520 --> 0:12:58.800 they possibly can without it accidentally blowing up in their face. 0:12:59.080 --> 0:13:03.120 Well, and this was also an era that was called 0:13:03.160 --> 0:13:05.960 sort of the open graph era of Facebook. Basically, they 0:13:05.960 --> 0:13:08.199 were trying to turn Facebook from an app where you 0:13:08.240 --> 0:13:11.960 would go and you know, stock your crushes from college 0:13:12.160 --> 0:13:18.120 or whatever, to a platform where other apps could build things. 0:13:18.160 --> 0:13:22.199 There were apps like Farmville that you could play inside Facebook. 0:13:22.240 --> 0:13:25.959 You could use Facebook to log into Spotify or any 0:13:26.040 --> 0:13:30.800 other number of services, and that was part of this 0:13:31.559 --> 0:13:34.400 quiz push was like you could build things on top 0:13:34.440 --> 0:13:37.840 of Facebook, and then Facebook would actually send some of 0:13:37.880 --> 0:13:41.920 that data to the developers of the quizzes. This was 0:13:41.960 --> 0:13:46.200 the same protocol, the same process that they used to 0:13:46.400 --> 0:13:49.520 connect people with Farmville, games and Spotify. 0:13:50.080 --> 0:13:53.760 The best case use of that data would just be 0:13:53.840 --> 0:13:57.840 for that app to reach more users or at this point, 0:13:57.920 --> 0:14:01.280 what's their reason for being just so give and take 0:14:01.400 --> 0:14:03.160 with this data. 0:14:03.280 --> 0:14:05.000 So Facebook used. 0:14:04.800 --> 0:14:09.199 To have this very permissive what they call an API 0:14:09.360 --> 0:14:12.360 and Application programming interface. It's essentially just a piece of 0:14:12.360 --> 0:14:14.880 software that other lets other pieces of software talk to 0:14:14.880 --> 0:14:18.520 each other. And there was a time when Facebook invested 0:14:18.559 --> 0:14:21.680 really heavily in that because they wanted essentially every other 0:14:21.840 --> 0:14:25.240 software to connect whatever they were building to Facebook. Facebook thought, 0:14:25.440 --> 0:14:27.200 this is the way that we are going to grow 0:14:27.240 --> 0:14:30.640 and take over the world is essentially everything runs through this. 0:14:31.160 --> 0:14:33.680 And the way that Facebook was able to attract so 0:14:33.760 --> 0:14:36.720 many of those people to its platforms was by saying 0:14:37.080 --> 0:14:39.680 we will give you data that you can't get anywhere else. Right, 0:14:39.720 --> 0:14:41.360 there aren't a lot of other places where you can 0:14:41.400 --> 0:14:45.320 go if you're a legitimate researcher, a shady researcher, somebody 0:14:45.360 --> 0:14:48.400 who's making a mobile game, and immediately get not just 0:14:48.560 --> 0:14:50.800 my name and my email address and my phone number, 0:14:50.880 --> 0:14:53.480 but also at the time, the names of all of 0:14:53.480 --> 0:14:56.240 my friends and their phone numbers and their email addresses. 0:14:56.440 --> 0:14:59.000 So for people like the ones who wound up building 0:14:59.040 --> 0:15:02.240 the app that led to Cambridge Analytica, this was a 0:15:02.320 --> 0:15:05.120 gold mind. It was an absolute bananza and it really 0:15:05.160 --> 0:15:07.040 benefited Facebook until it did. 0:15:07.280 --> 0:15:09.360 And do you think we've gotten just back to that 0:15:09.480 --> 0:15:13.920 initial question about how you know we share information online? 0:15:14.120 --> 0:15:17.000 Have we gotten any better at this or I mean, 0:15:17.000 --> 0:15:19.680 in casey, I love your analogy, Like if we get 0:15:19.680 --> 0:15:22.600 a discount on something, we're just like, yeah, take whatever 0:15:22.800 --> 0:15:26.080 whatever you want. Here's here's my mother's maiden name. But 0:15:27.360 --> 0:15:29.960 are we just that lazy, Like we're just our brains 0:15:29.960 --> 0:15:32.320 are just wired for heuristics, like we just want the 0:15:32.320 --> 0:15:33.480 shortcuts everywhere. 0:15:34.240 --> 0:15:38.160 I think there have been some major changes and you 0:15:38.200 --> 0:15:40.520 can see it when you log onto social media, right, 0:15:40.600 --> 0:15:43.120 you used to log onto Facebook and Instagram and you 0:15:43.160 --> 0:15:45.840 see your friends and family. Now you see a clip 0:15:45.840 --> 0:15:48.680 from a podcast you've never heard of, right, And that's 0:15:48.880 --> 0:15:50.880 you know, has several causes behind it, but one of 0:15:50.880 --> 0:15:53.520 the big ones is that Americans are sharing less. The 0:15:53.560 --> 0:15:56.120 whole bargain that they struck with Facebook of Hey, I'll 0:15:56.120 --> 0:15:58.920 give you all my information in exchange for you keeping 0:15:59.000 --> 0:16:01.480 up on the divorces of my high school classmates. That 0:16:01.560 --> 0:16:03.960 bargain doesn't feel like that's as good anymore. And so 0:16:04.000 --> 0:16:06.680 all the real sharing is happening in these private group chats, 0:16:06.720 --> 0:16:08.560 and all the public stuff is kind of more you know, 0:16:08.680 --> 0:16:11.360 creator content and you know, celebrities and all of that. 0:16:11.600 --> 0:16:13.640 So in that respect, I really do think things have 0:16:13.760 --> 0:16:16.840 changed now. At the same time, can I get you 0:16:16.920 --> 0:16:19.680 to give up information if I give you something of 0:16:19.880 --> 0:16:23.400 relatively low value? Yes, I think that is still broadly 0:16:23.440 --> 0:16:25.640 the case across you know, many different dimensions. 0:16:25.640 --> 0:16:29.960 Well, and I also think like AI has become the 0:16:30.000 --> 0:16:32.760 sort of next frontier of this because people are sharing 0:16:33.120 --> 0:16:35.720 all kinds of stuff which have spots. People are having 0:16:35.760 --> 0:16:40.480 therapy sessions. Yeah, they're confessing their crimes. They're they're like 0:16:40.800 --> 0:16:43.880 people are doing people are having these conversations that are 0:16:43.960 --> 0:16:47.400 very intimate in some cases, and I don't think they 0:16:48.240 --> 0:16:51.520 know or particularly care what's happening to that data on 0:16:51.560 --> 0:16:52.120 the other end. 0:16:52.720 --> 0:16:56.480 Wow, yeah, I don't know. I just I've gotten much 0:16:56.480 --> 0:16:59.680 more anxious about this over the last few years, kinds 0:16:59.680 --> 0:17:00.320 of I'm. 0:17:00.160 --> 0:17:00.880 Just doing it. 0:17:01.760 --> 0:17:04.520 Oh my gosh. Well, I mean the stuff I've already 0:17:04.520 --> 0:17:07.960 told to chat GBT, it's all. That's all just in 0:17:08.000 --> 0:17:17.920 the murder category. All right, let's get back into this story. 0:17:18.280 --> 0:17:21.719 So the app we're talking about this was collecting all 0:17:21.720 --> 0:17:24.040 this data and according to Facebook's rules, you're not allowed 0:17:24.040 --> 0:17:27.600 to repurpose this data in anyway. But Cogan was. He 0:17:27.760 --> 0:17:30.240 was selling this data to a little political consulting firm 0:17:30.240 --> 0:17:33.640 in the UK called Cambridge Analytica. In fact, it turns 0:17:33.640 --> 0:17:36.560 out Cambridge Analytica had actually paid Cogan to create the 0:17:36.600 --> 0:17:40.399 app in the first place. So what is Cambridge Analytica. 0:17:40.400 --> 0:17:43.080 Well founded in twenty thirteen, the company's mission was to 0:17:43.119 --> 0:17:46.760 take mountains of online data from potential voters and turn 0:17:46.800 --> 0:17:50.520 it into political leverage. The playbook borrowed a page from 0:17:50.520 --> 0:17:54.520 the US military's idea of psyops and applied it to elections, 0:17:54.960 --> 0:17:57.880 using targeted ads and so forth. They referred to this 0:17:57.960 --> 0:18:02.840 as quote, psychographic messaging. This is so Orwellian like it's 0:18:02.880 --> 0:18:06.560 it just sounds like big Brother hired an ad agency 0:18:07.760 --> 0:18:09.040 to come up with these terms. 0:18:09.200 --> 0:18:11.320 It does make me wonder whether if all of the 0:18:11.400 --> 0:18:14.600 things in this scandal had had like different names, it 0:18:14.600 --> 0:18:19.520 would have like Cambridge Analytica just sounds spooky. It's psychographic 0:18:19.560 --> 0:18:22.920 messaging just sounds spooky. It's like they have been called 0:18:23.000 --> 0:18:26.320 like political hot trends one oh one. 0:18:26.440 --> 0:18:28.159 Who wouldn't have been a scandal? I don't know. 0:18:28.960 --> 0:18:30.960 I'm not sure that one would have grabbed hold Kevin, 0:18:31.000 --> 0:18:35.679 but yeah, uh no. But you're right, and global science 0:18:35.720 --> 0:18:37.640 research is part of that too. 0:18:38.640 --> 0:18:40.720 And I think it's important to say, as we start 0:18:40.720 --> 0:18:43.840 to get into this looking back, I think many of 0:18:43.880 --> 0:18:45.880 the claims that were made by all of the players 0:18:45.920 --> 0:18:47.920 here were just wildly overstated. 0:18:48.080 --> 0:18:49.600 Right, it is both. 0:18:49.480 --> 0:18:53.760 True that Americans were very, very concerned about this psychographic 0:18:53.800 --> 0:18:57.160 targeting that was, you know, going on during election. It's 0:18:57.240 --> 0:18:59.720 also true that it almost certainly had nothing to do 0:18:59.800 --> 0:19:01.199 with the outcome of the election. 0:19:01.600 --> 0:19:04.480 Right, we'll get to that. We'll get to don't get 0:19:04.520 --> 0:19:06.640 ahead of me case. I'm sorry, geez. 0:19:06.760 --> 0:19:09.359 He's always doing this microphone. 0:19:09.440 --> 0:19:13.280 Okay, good to know. We'll just cut Casey out of 0:19:13.280 --> 0:19:16.159 this podcast. I think we can do that. We have 0:19:16.320 --> 0:19:20.320 the technology, or we'll just create an AI Casey that's 0:19:20.359 --> 0:19:24.680 way more compliant and just easier. Back to Cambridge Analytica. 0:19:24.720 --> 0:19:29.119 Running the show was CEO Alexander Nix. There was, also, 0:19:29.359 --> 0:19:31.760 of course, Steve Bannon, who you may know as Trump's 0:19:31.800 --> 0:19:36.520 former chief strategist and general architect of chaos. Bannon was 0:19:36.560 --> 0:19:40.320 one of the founders of Cambridge Analytica and later it's VP. 0:19:40.640 --> 0:19:44.760 Now bankrolling the whole operation were hedge fund billionaires Robert 0:19:44.800 --> 0:19:49.400 Mercer and his daughter Rebecca, longtime backers of conservative causes, 0:19:49.840 --> 0:19:52.800 they were also the founders and part owners of Breitbart News, 0:19:52.840 --> 0:19:57.000 where Bannon was executive chairman for several years. So, using 0:19:57.040 --> 0:19:59.280 the data from Cocanza app, Cambridge Analytica put together a 0:19:59.359 --> 0:20:03.679 robust se pychological profiles and personality scores for Facebook users. 0:20:04.080 --> 0:20:09.000 They then matched these profiles with US voter records, and voila. 0:20:09.200 --> 0:20:13.240 They now had a custom data set designed to predict, target, 0:20:13.280 --> 0:20:18.360 and potentially sway people's votes. Now, despite only about two 0:20:18.440 --> 0:20:21.879 hundred and seventy thousand people downloading this app and using it. 0:20:22.320 --> 0:20:27.000 The API loophole at Facebook allowed Cambridge Analytica to mine 0:20:27.119 --> 0:20:33.960 data from eighty seven million people. Holy shit, Holy shit. 0:20:34.880 --> 0:20:38.200 Who bears the responsibility here? Like? Is it the users 0:20:38.200 --> 0:20:41.160 for sharing the data? Is it the bad actors collecting 0:20:41.240 --> 0:20:44.359 and manipulating? Is it the platforms like or all of 0:20:44.520 --> 0:20:47.000 or is it just an holy toxic three way? 0:20:47.720 --> 0:20:50.520 I think it is a bit of a toxic three way, 0:20:50.800 --> 0:20:53.480 you know, Like I imagine that you ed and most 0:20:53.520 --> 0:20:56.359 of your listeners have had the experience of you install 0:20:56.400 --> 0:20:58.879 an app on your phone and it asks you, do 0:20:58.920 --> 0:21:04.200 you want to share your contacts with this app? For developers, 0:21:04.280 --> 0:21:08.240 that can be a very effective way of finding other customers, 0:21:08.320 --> 0:21:10.080 or if you're trying to grow a social network, that 0:21:10.160 --> 0:21:12.399 can help to grow a social network. And if you 0:21:12.400 --> 0:21:14.560 think that some social networks are good, maybe there is 0:21:14.600 --> 0:21:17.120 some value there. But it's also true like if you're 0:21:17.280 --> 0:21:20.000 one of the people whose information is being shared, like 0:21:20.080 --> 0:21:22.880 arguably it's kind of being shared without your consent by 0:21:22.880 --> 0:21:24.679 like one of your friends, and it's being used in 0:21:24.720 --> 0:21:28.000 ways that you have no real control over. And so 0:21:28.520 --> 0:21:31.119 I do think that was bad and it is why 0:21:31.359 --> 0:21:35.040 you know, that practice has been dramatically curtailed in more 0:21:35.080 --> 0:21:35.760