How Much Does Google Know About Me? | Seth Stephens-Davidowitz | Modern Wisdom 134

1. 0:00 – 1:59

   From ad-click optimization to “data for social good”

   1. SSSeth Stephens-Davidowitz0:00

      There are lots of people in this area and, you know, data science is just exploding in all kinds of ways and I think a lot of people are go- definitely millennials or people younger than millennials are also looking ... You know, it seems like the values are shifting a little bit where it's less about just making money. So I think initially everyone's kind of like, "Oh, data science, that's a lucrative field. I can get a job at-"

   2. NANarrator0:19

      Mm-hmm.

   3. SSSeth Stephens-Davidowitz0:20

      ... you know, getting people to click on ads or, uh, work, get a job in finance." Which, which is tot- are totally fine jobs. They're kind of bored of studying. You know, they- they- they like data science but they're kind of bored of s- of getting people to click on ads and they're, you know, they, they feel kind of unfulfilled and lacking purpose and I think there are ways to use this data, uh, towards social good as well.

   4. CWChris Williamson0:43

      (wind blows) Seth. Hi, man. How are you doing?

   5. SSSeth Stephens-Davidowitz0:45

      Really good. How are you?

   6. CWChris Williamson0:47

      Yeah. Very good, thank you. I'm excited for today. Big data and all this sort of stuff. It's going to be cool.

   7. SSSeth Stephens-Davidowitz0:55

      Yeah. I hope so.

   8. CWChris Williamson0:56

      Looking forward to it. So first things first, how do you describe what you do for work or on a day-to-day basis?

   9. SSSeth Stephens-Davidowitz1:05

      Uh, well, so I guess I'm, I guess I'm a data scientist and an- and an author, writer. So, uh, you know, I, I spend most of my time on books. I wrote Everybody Lies. I'm trying to write ... I'm in the process of writing another book. And, uh, other than that, I don't know. A lot of random projects. I do random consulting for companies and, uh, there's not necessarily a standard day, but-

   10. CWChris Williamson1:30

       I got you. Is it a lot of time on spreadsheets or similar sort of applications?

   11. SSSeth Stephens-Davidowitz1:34

       Yeah. Uh, R, the coding language. There's a lot of time on that, kind of moving back and forth between R and, uh, Google Docs, uh, 'cause I guess the combination of data science, which is R, and writing, which is Google Docs.

   12. CWChris Williamson1:46

       Got you.

   13. SSSeth Stephens-Davidowitz1:46

       Uh, and then sometimes just researching. A- a lot of, like, reading, so a lot of reading other people's studies. Kind of reading what, uh, other people are talking about since I can't just, uh, write about my own studies.

   14. CWChris Williamson1:57

       Which is a shame.

   15. SSSeth Stephens-Davidowitz1:58

       Yeah. (laughs)

2. 1:59 – 5:43

   The “How Big Is My Penis?” story: what private searches reveal

   1. CWChris Williamson1:59

      Yeah. It'd be lovely. Um, so you wrote a book called Everybody Lies, New York Times bestseller. But I heard that you wanted to call it How Big Is My Penis? Is that right?

   2. SSSeth Stephens-Davidowitz2:11

      (laughs) Tha- tha- that- that is, that is correct, yeah. I-

   3. CWChris Williamson2:13

      You wanted to call your book How Big Is Your Penis?

   4. SSSeth Stephens-Davidowitz2:17

      No, How Big Is My Penis? Um-

   5. CWChris Williamson2:18

      How Big Is My Penis? N- yes.

   6. SSSeth Stephens-Davidowitz2:19

      Well, no. Basically the reason for it is that that's one of the top ... Well, I, I talk about how men ask more, Google more questions about their penis than any other body part. (laughs) And that one of the top questions they ask Google about their penis is, "How big is my penis?" Which is just, like, an absurd question to ask Google.

   7. CWChris Williamson2:36

      (laughs)

   8. SSSeth Stephens-Davidowitz2:36

      Right? They're not going to be able to answer that. So I ki- I kind of thought that that title would kind of get the flavor of what some of the, some of the things I was talking about, uh, that kind of ... But I think one of the things that my research has shown is the absurdity of the human condition, the absurdity of people. We kind of, people kind of put on a very presentable front, but in the privacy of their own home on their s- Google search engine or the websites they visit and they go to Pornhub they kind of just show, uh, a different side of themselves which is a little bit stupider, a little bit less polished, a little bit weirder, a little bit sometimes nastier. Uh, but I think, uh, it's kind, it's kind of, um, an interesting view of people that we haven't really previously had.

   9. CWChris Williamson3:16

      Mm. Yeah, that totally unfettered view where you, you don't think that anyone else is watching, but the data analysts are watching. You are.

   10. SSSeth Stephens-Davidowitz3:25

       Well, th- the data I analyze is all a- anonymous and aggregate, so I don't know that any particular man searched how big is my penis. I just know that lots of men apparently-

   11. CWChris Williamson3:33

       (laughs)

   12. SSSeth Stephens-Davidowitz3:35

       ... uh ...

   13. NANarrator3:37

       (laughs)

   14. CWChris Williamson3:37

       You know what I think? I think there is an opening in the market for an app which can use AR, like augmented reality. You hold it up to your face, you angle the penis, and then it, it works out how big it is.

   15. SSSeth Stephens-Davidowitz3:52

       Well, how would ... You'd have to know how big the face is. Right? You'd need, like, something behind. You'd need to be, like, standing in front of the Eiffel Tower or something. Something-

   16. CWChris Williamson4:01

       Uh-huh. But then you'd need to know the distance.

   17. SSSeth Stephens-Davidowitz4:03

       And then know the distance. Uh, yeah, but I don't think just com- Yeah.

   18. CWChris Williamson4:07

       Mm.

   19. SSSeth Stephens-Davidowitz4:07

       I don't know.

   20. CWChris Williamson4:08

       Yeah, a ruler's easier, isn't it? Anyway, anyway, we're getting off, we're getting off track. Um ...

   21. SSSeth Stephens-Davidowitz4:13

       (laughs)

   22. CWChris Williamson4:13

       (laughs) Uh, so but you have this unique-

   23. SSSeth Stephens-Davidowitz4:16

       I think sometimes a lot of men, like, when they ... They, the question is so important to them that I think one of the reasons they ask that is, like, there's almost like men don't want to face the moment of truth. Like, it would be easy to just measure, but they're like, "Is there another way to find the answer?" 'Cause there's so much insecurity, I guess, and, uh, fear and anxiety around the process. Um-

   24. CWChris Williamson4:34

       Have you seen the South Park episode where they do that?

   25. SSSeth Stephens-Davidowitz4:37

       Where they do what?

   26. CWChris Williamson4:38

       They measure their penis. Have you seen this?

   27. SSSeth Stephens-Davidowitz4:40

       No, I haven't seen it.

   28. CWChris Williamson4:41

       Anyone who's listening, if you get the opportunity to go back and watch the South Park episode where they all measure each other's penises, it's unbelievable because they ... There's some people who have really small measurements and they all get really, really angry, so then they create this new formula where they adjust for the yaw, like the height of the penis, and then they divide it and they create this ratio, and it actually makes it, like, several feet long. And then every, all of the problems that everybody had has been completely accounted for.

   29. SSSeth Stephens-Davidowitz5:07

       That used to happen in sports. I remember, like, uh, my friends and I, we were nerdy but we were, we were like, we were nerdy and very, very competitive. So, like, we'd get out the stats at the end of the year. I played baseball. And then, like, we'd all, like, create the ... You know, you, you'd have, like, the basic stats like your batting average, your home runs, your RBIs, your runs, whatever, and then we'd all create metrics, like how we sh- we should weight the different s- categories-

   30. CWChris Williamson5:30

       Mm-hmm.
3. 5:43 – 8:41

   Why the book is called Everybody Lies: surveys vs. behavioral data

   1. CWChris Williamson5:43

      So, um, why, why call it Everybody Lies then? Why, why was the book Everybody Lies?

   2. SSSeth Stephens-Davidowitz5:47

      Well, I don't ... Yeah. First of all e- my publisher chose that title, but I, I, I think it's a good title. It kind of gets to the point. Some of my research, which is that people in their day-to-day life, uh, lie a lot. So we lie face-to-face. We also lie, surprisingly somewhat, to anonymous surveys. That's the tr- traditional way we tried to understand human beings historically-... uh, what they're thinking, what they're going to do, what they have done, why they do the things they do is surveys. So Gallup or Pew or Quinnipiac or another organization will ask people questions. And even with this methodology, even though it's anonymous, uh, people do shade their answers to try to sound good. Maybe they're lying to themselves. Uh, maybe they're just in a habit of lying. They don't have an incentive to tell the truth, so they are more likely to say they voted in a previous election than they d- they did. They exaggerate how frequently they have sex if you ask them in surveys. You compare that to actual condom sales, like I show in the book, like, it doesn't correspond at all. Uh, so, you know, how fre- uh, wh- uh, wh- what, what media they consume, how frequently they're, uh, watching kind of high-brow stuff versus low-brow stuff, they'll all lie in these surveys. But now thanks to the internet and largely Google but other websites as well, we have kind of a new window into people that I think gets ... cut through the lies a lot. So Google, for example, you see people, uh, searching how to vote, where to vote. That's really predictive what they're actually going to vote. You talk ... See them searching about sexless marriage or sexless relationship, or my boyfriend doesn't want to have sex with me, uh, and then you get a ... kind of a more realistic picture how men ... what happens in, in many relationships. And, uh, you a- see people s- uh, searching for their sexual insecurity, or you see people, uh, what, what they actually want to read, the media they actually want to consume. I think it's really much more accurate. Uh, you know, and then there's a whole section on pornography. I mean, talk about an area where there's been, uh, a huge amount of deception because it's kind of there are so many taboos in that area. I think we really do have unprecedented window into people's kind of sexual, sexuality-

   3. CWChris Williamson7:47

      Mm.

   4. SSSeth Stephens-Davidowitz7:47

      ... now that we've never had before thanks to pornography. And I analyze in, in Everybody Lies, uh, kind of I have a whole section where I analyze just Pornhub, what people actually watch on Pornhub, uh, which-

   5. CWChris Williamson7:57

      Yeah.

   6. SSSeth Stephens-Davidowitz7:57

      ... uh, make people giggle, but it's also, I think, pretty fascinating and re- and kind of revolutionary, kind of this window into, uh, people's minds that we didn't have. Uh, I mean, you know, you ... All of human history up until five years ago, we didn't know that. We didn't know what people fantasized about sexually. We might have had some clues by, uh, you know, novels people wrote or some theories that some, uh, you know, people like Freud came up with, or, you know, there are some surveys, the Kinsey study, but it's kind of a little bit, you know, there's question on how accurate that was, uh, kind of a, uh, a biased sample. But I think now we kind of have really, uh, much better, uh, clues in that arena. So, uh, kind of over and over again, I think, uh, we're ... So, so there are just so many areas where I think we're ... we do have unprecedented, uh, insights into people.

4. 8:41 – 10:34

   The art of data science: creativity, pattern-finding, and cultural oddities

   1. CWChris Williamson8:41

      Yeah. And then it's your job to sift through this aggregated anonymous data and try and analyze or come up with some correlations between things and some interesting insights.

   2. SSSeth Stephens-Davidowitz8:53

      Yeah. Uh, that, yeah, I think, uh, that's kind of the talent. There's kind of an art to data science. I think data science people kind of don't usually associate with creativity. Uh, we usually think of art, you know, a painter's creative, a musici- musician's creative. A data scientist we think is kind of this nerdy, uh, just kind of sit by a computer, plugs the numbers in. And, uh, I think creativity is a huge part of data science. Otherwise, you kind of just drown in the data and you don't really know what's interesting. So kind of, uh, the creativity is kind of looking through all that and finding those nuggets of wisdom, you know, the how big is my penis search or the-

   3. CWChris Williamson9:25

      Mm-hmm.

   4. SSSeth Stephens-Davidowitz9:26

      ... most ... Uh, the other ... My, my other favorite is Indian, um, men. The top Google search my husband wants in India is my husband wants me to breastfeed him. (laughs) And that's like India-

   5. CWChris Williamson9:37

      You're kidding.

   6. SSSeth Stephens-Davidowitz9:37

      Yeah. No. I'm not joking. That's, like, India and nowhere else. And also, breastfeeding porn is, uh, reasonably popular in India and nowhere else.

   7. CWChris Williamson9:44

      (laughs)

   8. SSSeth Stephens-Davidowitz9:44

      Uh, inaudible ] breastfeeding porn, and that's kind of the types of things that ... I mean, that's kind of, like, that's kinda ... That kinda ... That's kinda shocking in many ways 'cause it's, it's not talked about. It wasn't really acknowledged without this data, and it kind of shows that some sort of, uh, fetish can develop in one part of the world and nowhere else, uh, and become kind of somewhat widespread. I don't think it's, you know, the majority of Indian men, but, uh, uh, cer- certainly a large number.

   9. CWChris Williamson10:07

      A significant minority, yeah.

   10. SSSeth Stephens-Davidowitz10:09

       Significant minority, without being talked about at all. Uh, like something you care-

   11. CWChris Williamson10:12

       And how else, how else would you find about that, right? How else would you know?

   12. SSSeth Stephens-Davidowitz10:15

       Yeah. Uh, and you won't find out about it without, uh, without this type of data. So that, yeah, so things like that, I mean, you know, kind of finding those insights amidst, you know, just kind of rows and rows of data, uh, is kind of a, an art and something that you get better at, uh, with practice and takes a lot of creativity. Uh, but yeah.

5. 10:34 – 15:01

   Getting Pornhub data—and why academia underuses it

   1. CWChris Williamson10:34

      One, one of the things that I absolutely love, Pornhub issue a bunch of stats at the end of each year. Nothing-

   2. SSSeth Stephens-Davidowitz10:40

      Yeah.

   3. CWChris Williamson10:40

      They don't delve into it like you do, but it's like amount of playtime, uh, the top searches by area, and, and stuff like that. And even that, (clears throat) even the things that they-

   4. SSSeth Stephens-Davidowitz10:49

      Yeah.

   5. CWChris Williamson10:49

      ... do in-house is pretty, pretty powerful.

   6. SSSeth Stephens-Davidowitz10:52

      Oh, yeah. And I know some of the data scientists there, and they're, they're really good. And they're ... And, you know, and it is just, it's kinda, I kinda, you know, I think people don't take it almost seriously enough, you know, and academia or sociologists. Uh, I kind of reached out to Pornhub when I was writing my book, and I'm like, uh, "Can I, I really want to look at your data." I'm ... I explained I'm a data scientist. I showed them some of the work I had done writing columns for New York Times. And, uh, they, they agreed to give me the data. But I, I would've thought that ... I'm like, "Do you get 10 emails a day from, like, sex researchers or sociologists?"

   7. CWChris Williamson11:22

      Mm-hmm. Mm-hmm.

   8. SSSeth Stephens-Davidowitz11:22

      And, and they were, like, kind of like, "No," you know. You know, they're kind of comfortable with the methodologies they've been using, uh, for 50, 60 years. And, uh, you know, the ... it ... academia kind of moves slowly and doesn't change, uh, that, that quickly. And, uh, you know, so it's kind of ... I think, I think your instinct and my instinct is all, uh, is, like, wow, that's fascinating. Like, you know, and, and, uh, you know, relative to kind of just another survey about human beings and what they're doing, what they're interested in sexually, actually seeing, uh, kind of, you know, this data from, uh, this enormous site of, of, you know, sexual fantasy or sexual desire is pretty, pretty wild and, you know, pretty f-

   9. CWChris Williamson12:02

      What ha- what, what happens when you receive the Pornhub stats in your Dropbox or, like, over WeTransfer or whatever it is? Is it just like this-... gargantuan, kinda sticky, like, disgusting (laughs) sort of bit of-

   10. SSSeth Stephens-Davidowitz12:15

       Uh, uh-

   11. CWChris Williamson12:15

       ... stats or whatever? Did it feel-

   12. SSSeth Stephens-Davidowitz12:17

       I don't know. I mean, I-

   13. CWChris Williamson12:17

       ... icky opening that file versus opening another one?

   14. SSSeth Stephens-Davidowitz12:19

       I'm a little bit, like, uh, socially oblivious. So, like, for me, it's just like I might as well have just been getting a data set of, like, interest rates historically, uh, or, you know, and inflation historically and running the numbers on that. To me, I'm just... Except this one was more interesting to me.

   15. CWChris Williamson12:32

       Mm-hmm.

   16. SSSeth Stephens-Davidowitz12:33

       But I'm not like, "Oh, this is a weird use of my time," or, like, a lot of people would be kind of queasy looking at this or embarrassing, embarrassed looking at this. I don't really... Whatever part of the brain normal people have, uh, that kind of shi- shies away from that type of research or that type of activity, I definitely-

   17. CWChris Williamson12:49

       You've-

   18. SSSeth Stephens-Davidowitz12:49

       ... uh, done so. I, I wasn't at all-

   19. CWChris Williamson12:51

       You've found s- found the right industry to be in, then?

   20. SSSeth Stephens-Davidowitz12:54

       Yeah. Well, I kind of just invented (laughs) the...

   21. CWChris Williamson12:56

       (laughs) Yeah, okay.

   22. SSSeth Stephens-Davidowitz12:57

       ... the idea of this industry, I think. But yeah, it does fit my personality, yeah. If you have, like... Uh, you're kind of a little bit, I guess, shameless and, uh, not queasy. And I think, uh, there's kind of a good combination, which is that I built up... Before I wrote my book, I built up all these credentials. I did a PhD in economics at Harvard, and I went to Stanford s- like philosophy. I was a writer for The New York Times. I worked at Google. So, I think when I do this, people aren't like... People are like, give me, like, a lot of leeway. Like, if I was, like, some 20-year-old in my bate- uh, like, 20-year-old, like, with no, no qualification, I'm just like, "Hey, you know, I'm analyzing porn," everyone would be-

   23. CWChris Williamson13:32

       (laughs)

   24. SSSeth Stephens-Davidowitz13:32

       ... like, you know, "That pervert." Like, "What's (laughs) , what's that guy doing?" But, like, for... With me, it's like, "Oh, you know, this is science." Like, I, I think there, there... People assume a certain respectability with me just based on the credentials that I kind of get away, uh, with, uh, you know, more than I, than I otherwise would. I think, you know, if, if, if I didn't have the credentials and I came out with the research I did, I think a lot of people would be like, "Wait, that guy just locked himself in his apartment and watched porn for two years." (laughs)

   25. CWChris Williamson13:59

       (laughs) Yeah. This is just him. This is him and his breastfeeding, his breastfeeding-

   26. SSSeth Stephens-Davidowitz14:03

       Yeah.

   27. CWChris Williamson14:03

       ... stuff. Yeah.

   28. SSSeth Stephens-Davidowitz14:04

       But, uh, but, you know, with, with, with the credentials, it's kind of like, "Oh, that guy is, you know, exploring the deepest, uh, recesses of the human psyche."

   29. CWChris Williamson14:11

       Mm-hmm. He's got this veneer of, like, respectability-

   30. SSSeth Stephens-Davidowitz14:14

       Yeah.
6. 15:01 – 19:41

   What porn data shows: women’s preferences, universality, and “why” questions

   1. CWChris Williamson15:01

      Got you. So what else, what else have we learned from porn? Let's stay with porn for a second before we get away from porn. Were there any other s- surprising things that came up other than

   2. NANarrator15:08

      You know?

   3. SSSeth Stephens-Davidowitz15:08

      There's some of them... Some of them, I think, when I ca-... When the book came out, people said they were really shocking. To me, I didn't find it shocking. So for example, the popularity of rape porn among women is very, very striking in data. A huge percent of heterosexual porn searches or vide- videos watched by women are for kind of violent humiliation, uh, against a woman.

   4. CWChris Williamson15:30

      Mm-hmm.

   5. SSSeth Stephens-Davidowitz15:30

      And much more popular among women than men. About twice as popular among women than men. Uh, that, that didn't shock me. There have been surveys that kind of have talked about that a lot of women have these types of fantasies. And kind of just in my conversations with friends, I have very honest friends, this is kind of... You know, it's kinda come up in my life that-

   6. CWChris Williamson15:52

      Mm-hmm.

   7. SSSeth Stephens-Davidowitz15:52

      ... that it, that it didn't... When I, when I saw the data, I'm not like, "Oh, my God. That shocked-"

   8. CWChris Williamson15:56

      Mm-hmm.

   9. SSSeth Stephens-Davidowitz15:56

      But I kind of looked at it and...

   10. CWChris Williamson15:57

       I mean, loo- look at, look at the most popular book of the last decade.

   11. SSSeth Stephens-Davidowitz16:02

       Yeah. 50 Sh- Right, yeah. Yeah, so-

   12. CWChris Williamson16:03

       50 Shades of Grey.

   13. SSSeth Stephens-Davidowitz16:04

       So... Right. But, you know, it still is, it still did have... I think what was more interesting is you actually can compare the percentages around the world, and it's not correlated at all with how women are treated. That's kind of interesting. That's not... It's not like also-

   14. CWChris Williamson16:17

       How, how so? What do you mean?

   15. SSSeth Stephens-Davidowitz16:18

       In other words, so you have, like, some areas like Berkeley, California or, like, parts of, you know, or parts of, uh, you know, or you have Sweden or Finland or Netherlands, which Finland now has a female prime, prime minister and, like, there, where there is really much progressive attitudes towards women. Then you have areas like, you know, Saudi Arabia or Iran where women are really held back. And it's not like... So you could really rank kind of, you know, how, how women are treated. Are they re-... Are they treated like second-class citizens?

   16. CWChris Williamson16:44

       How egalitarian it is.

   17. SSSeth Stephens-Davidowitz16:46

       Yeah. And you, you might... You, you could imagine that that could affect how women think about themselves and kind of, you know, affect their fantasies, but it doesn't seem to... You know, it's kind of a pretty universal-

   18. CWChris Williamson16:57

       It's this universal-

   19. SSSeth Stephens-Davidowitz16:57

       Yeah. Or-

   20. CWChris Williamson16:57

       ... desire to watch some, like, hardcore porn.

   21. SSSeth Stephens-Davidowitz16:59

       Even if, even if women are growing up telling, you know, saying they could be everything, they're, they're equal to men, they're, you know, there, there still is, uh, that fantasy seems to exist in about the same number as the places where women are saying, like, "Men should dominate you. Men are the, you know, the, the lead- should be the leaders of society." So that was pretty interesting.

   22. CWChris Williamson17:16

       That is, that is really, really fascinating. Do you ever ask the why question? Do you ever bother to delve into that, or do you just sort of stick to the, to the data?

   23. SSSeth Stephens-Davidowitz17:23

       So I try to in the next, in the next level. You know, I think there are... Uh, you know, it's, it's... I think the initial thing... Initially, I was kind of just presenting the facts. And I have, I have made much progress. I, I, I thought other people would come up with theories. So I thought when I wrote the breastfeeding one, people are gonna like... Maybe there's some explanation that's very obvious that, you know, I didn't know that... But nothing's come up, uh, you know, to kind of explain this. So I, I don't know. It's, it's, it's tough. Uh, I think... You know, I think there are definitely some areas. So definitely I think people are more attracted to people they grew up around, uh, or like, people like them, I think p- probably maybe more... Maybe because of their environment. So there's kind of there's been this long idea that opposites attract. Uh, and I think if you see in dating data, it's not true at all. So in dating data, it's very, very clear that people are drawn to people who are similar to themselves on just about any dimension you could measure.... and pornography data, it also seems to be the case. So if you look at areas that have high African American populations, uh, they tend to watch porn featuring African American ... Well, there's some men in that area, those areas tend to, uh, watch porn featuring African American women in large numbers. Uh, so it's not like, you know, uh, yeah, I think you could imagine that it would go the opposite way and that the fantasy of the Black man would be, you know, the, the white cheerleader or something. I don't know, that you could imag- uh, but it, it's definitely not the case. Uh, so that, that, that kind of, I think, also moves towards the why direction that maybe sexual ... I, I do think that one of the things you see is that, uh, sexual fantasies seems to be related to childhood in various ways. Uh, so people, uh, tend to have fantasies from their childhood. They fantasize about, uh, babysitters or, like, teachers or I think there's kind of a, there's kind of a maybe kind of key moment in childhood where people kind of get imprinted sexually that also, uh, dovetails with some other research in, uh, in fields of sexuality, uh, that childhood imprinting is really important in developing, uh, adult sexuality.

   24. CWChris Williamson19:27

       Mm-hmm.

   25. SSSeth Stephens-Davidowitz19:27

       So that would explain why if, if, you know, a, a Black man had grown up around a lot of Black women, uh, he'd probably be more likely to be attracted to Black women, uh, than a, than one who hadn't been grown up. Yeah.

   26. CWChris Williamson19:40

       Yeah. I get that.

7. 19:41 – 22:58

   Sexual orientation signals: lesbian porn among straight women and closeted men

   1. SSSeth Stephens-Davidowitz19:41

      Yeah.

   2. CWChris Williamson19:41

      What about, um, when we're talking about the split between heterosexual and homosexual, there must be some interesting insights there about how many men and women are being truthful about their sexuality?

   3. SSSeth Stephens-Davidowitz19:52

      Yeah. So one of the things that, that's also striking today, which also didn't surprise me just 'cause it had come up in conversations I've had with very honest friends is the popularity of lesbian porn among women who otherwise consider themselves totally straight. And I think, you know, I, like, uh, I don't think that they're necessarily in the closet. Uh, so, like, I think about 20 ... I think it was ... I, I don't remember these exact numbers. I think it's something like 20% to 25% of pornography, uh, views from women are for lesbian, explicitly lesbian porn. And I think, you know, uh, this will come up in surveys or focus groups, a lot of women, you know, that they live in Berkeley or San Francisco or areas where it's totally, uh, you know, or l- or lo- where, where, you know, in this day and age, I think it's pretty okay, uh, to be lesbian. I don't think there are many social pressures to be heterosexual. And they consider themselves straight. They only want to pursue relationships with men, but they're just like, "You know, the female body's hotter." Like, it's just they, they can, I think, disconnect kind of the real emotional... Women are maybe sometimes ... I don't want to over-generalize, but women are maybe better at disconnecting, uh, the kind of emotional, romantic connection to just like-

   4. CWChris Williamson21:03

      Mm-hmm.

   5. SSSeth Stephens-Davidowitz21:04

      ... your, your physical, uh, in, in interesting ways possibly. I don't know. It's, it's-

   6. CWChris Williamson21:08

      That's quite, that's quite unique insight, because I doubt that the equivalent would be true for men with g-

   7. SSSeth Stephens-Davidowitz21:13

      Yeah, so what-

   8. CWChris Williamson21:14

      ... with gay porn.

   9. SSSeth Stephens-Davidowitz21:14

      Yeah. What you see with men is that gay porn is about 5% of, you know, depending on ... It depends on the site you look at, but it seems like about 5% of male porn, maybe a little bit lower, is for gay porn.

   10. CWChris Williamson21:24

       Mm-hmm.

   11. SSSeth Stephens-Davidowitz21:26

       And one of the things you see is that it's a l- it's, it seems pretty clear, you know, it's almost as high in areas where it's hard to be gay, where a lot fewer men say they're gay. So places like Mississippi or Alabama or Tennessee, if you ask men, "Are you gay?" Only about 1% to 2% of men will say they're gay. Whereas in New York and California, about 4% of men will say they're gay. But the gay porn searches are almost the same everywhere, which I think says that, that a lot of the men in Mississippi are, are in, are in the closet. Uh, you know, I, and I, I do think you also see with the searches, so the number one search, uh, nec- like, one of the top searches right in the, the same session where someone searches gay porn is gay test. There's also kind of ridiculous searches.

   12. CWChris Williamson22:09

       (laughs)

   13. SSSeth Stephens-Davidowitz22:09

       I like the, "How big is my penis?" search. It's like they-

   14. CWChris Williamson22:11

       (laughs)

   15. SSSeth Stephens-Davidowitz22:12

       They ask you these ridiculous questions to try to figure out, like, are you gay. But I think, uh ... And these searches are much more common in, in the place where it's hard to be gay-

   16. CWChris Williamson22:21

       Mm-hmm.

   17. SSSeth Stephens-Davidowitz22:22

       ... are the gay test searches. And I think when you see these, these kind of, the search strings, uh, you kind of see a tortured man or man who's really uncomfortable with his sexuality and, and, and is trying to, is like trying to find some way to prove that he's not gay, uh, you know, by-

   18. CWChris Williamson22:36

       Almost to himself. He's not proving it to anybody else, right?

   19. SSSeth Stephens-Davidowitz22:39

       Exactly. Yeah, to some, some stupid internet test. So I think you do see kind of the torture that, uh, s- you know, it, that some g- gay men feel in places, uh, where it is hard to be gay, which is, you know, it's changing. Thankfully fewer and fewer parts of the world now, uh, don't have the anti-gay attitudes-

   20. CWChris Williamson22:57

       Mm-hmm.

   21. SSSeth Stephens-Davidowitz22:57

       ... but there still are a lot of places.

8. 22:58 – 25:24

   Search strings as narratives: suicide, stigma, and the herpes insight

   1. CWChris Williamson22:58

      Yeah. It's interesting now where you're talking about these strings of search terms where you can weave together a narrative of what's going on. Were there any other strings of search terms that you thought were particularly interesting or anything you've come across recently?

   2. SSSeth Stephens-Davidowitz23:11

      Yeah. So one of them I've been ... I'm working on an article on this now, but I've been looking at, you know, a dark topic with suicide. Uh, I think it's an important topic because you could imagine if you had a search string, if you had a string of searches and someone searches how to commit suicide-

   3. CWChris Williamson23:25

      Mm-hmm.

   4. SSSeth Stephens-Davidowitz23:26

      ... you could imagine looking at the searches from earlier and kind of getting a view of what's causing them to have that thought. I think that's ano- another area where we don't have right now, uh, a great idea of why everybody ... You know, why people think of suicide 'cause there's so much stigma around that whole area. And one ... Like, one of the things I found which is really surprising was a, a, a, a com- a somewhat common cause of suicidal thought among young people is herpes, getting diagnosed with herpes, the STD, uh, which kind of sounds a little ridiculous. Uh, it's not a, uh ... It's a disease that's actually pretty common, uh, and it's, it's not one with ... You know, it's not life-threatening. The physical symptoms are, to the best of my knowledge, uh, fairly limited. Uh, but I think the stigma is immense. And you can imagine a yo- a young person who's just gotten diagnosed with herpes, when, when people are 18, 19 years old, I mean, I, if I go back to my own, uh, experience in that age group-... you don't know what's going on. You don't know how the world works. It's such an, a time of extreme paranoia-

   5. CWChris Williamson24:30

      (sniffs)

   6. SSSeth Stephens-Davidowitz24:31

      ... uh, that there's something fundamentally wrong with you. I mean, it's kind of like the kids in Mississippi who were searching gay porn and gay tests. There's ... It's so easy to get paranoid, uh, that you are kind of, you know, uh, lacking in some fu- fundamental way, or, you know, you, you, that, uh, that, that you, n- you don't know what's common, what isn't common. I've, I've, I've heard stories of, you know, women that freak, little girls who freak out when they first grow breasts that they have breast cancer or something. Like, you know, it's like when, when you're a kid and when you're ... I think the po- possibility for, uh, kind of extreme paranoia is very high, and it can ... because the data says it can lead you to be s- in so much pain and so much anxiety that you actually, uh, think of committing suicide. Uh, so I think that's really, uh, a, a promising area of research to kind of get in the mindset of the suicidal, particularly young people-

   7. CWChris Williamson25:24

      Mm-hmm.

9. 25:24 – 29:43

   Role-model searches: “celebrities with herpes” and what people are really seeking

   1. SSSeth Stephens-Davidowitz25:24

      ... who've been feeling suicidal and maybe starting to fight these attitudes. The ot- another thing I found is the number one search when p- in the search string when people search herpes and how to commit suicide is celebrities with herpes, which is they're basically looking for role models. In fact, they ... That's actually a common search for just about any illness, so people who have depression search celebrities with depression. People who have back pain search celebrities with back pain.

   2. CWChris Williamson25:47

      Mm-hmm.

   3. SSSeth Stephens-Davidowitz25:47

      People are looking for role models or heroes of theirs who share their condition to, so that they feel less stigmatized. And I looked on Google what kind of happens when you search celebrities with herpes, and if you search celebrities with just about any illness, celebrities with depression, huge number of celebrities openly discuss their depression in part to fight the stigma. Celebrities with, with back pain, celebrities with an- and just about any illness you can think of, a long list of celebrities, uh, saying they're, uh, se- uh, kind of, uh, coming out of that, that closet, uh-

   4. CWChris Williamson26:18

      Mm-hmm.

   5. SSSeth Stephens-Davidowitz26:18

      ... to, to help. It ... I think, uh, uh, to their credit, to help, uh, their fans, uh, who might also be struggling, uh, with this problem. But celebrities with herpes, you see al- basically what the top sites, at least last time I checked was just celebrities denying they had this. Uh, you know-

   6. CWChris Williamson26:34

      (laughs) It's the opposite.

   7. SSSeth Stephens-Davidowitz26:36

      Yeah, then you can imagine ... All right. Like, put all this data together. You have young people. They've, they were diagnosed with herpes. They're thinking of committing suicide. They're looking for celebrities with herpes, and instead of greet, you know, getting greeted with a list of celebrities' videos saying, "You know, this is nothing to be ashamed of. It's not a big deal. I live with this." You know, it's ... Uh, you have celebrities saying, "I would never have-"

   8. CWChris Williamson26:56

      (laughs)

   9. SSSeth Stephens-Davidowitz26:56

      "... you know, an illness. It's uh ..." Uh, you know, yeah. It's, it's kinda like ... It kinda ... E- every time I tell this, it does make people giggle, and, you know, I admit I've giggled sometimes myself, but I think it actually is, like, kind of profound and, and serious and shows the power of this-

   10. CWChris Williamson27:09

       I don't know. So-

   11. SSSeth Stephens-Davidowitz27:10

       ... kind of this unvarnished window into the human mind, uh, and how powerful that could be, and, uh-

   12. CWChris Williamson27:15

       That-

   13. SSSeth Stephens-Davidowitz27:15

       ... you know that a lot of people, a lot of people are suffering in ways that are, that's, uh ... It makes us giggle 'cause they're so silly, but to that person, it's not silly.

   14. CWChris Williamson27:25

       Mm-hmm.

   15. SSSeth Stephens-Davidowitz27:25

       It's like they might as well be in Auschwitz. Like (laughs) i- they, they, they might, uh, like, eh, the pain that someone, a suicidal person is feeling, no matter what the cause of it, uh, is extreme. So, uh, you know, like, so I think getting a sense of-

   16. CWChris Williamson27:38

       (alarm blares)

   17. SSSeth Stephens-Davidowitz27:38

       ... getting a, getting a sense of, of, of, a better sense of, uh, the, the, the, uh, people's kind of darker thoughts, uh, can be, uh, really helpful and can, can ... You know, th- that kind of ... We kind of have a ... That kind of suggests kind of an obvious way to change it, get more celebrities to say (laughs) to cut, you know, say they have herpes and, and fight it-

   18. CWChris Williamson27:57

       I'm not sure ... I'm not quite sure how you mandate for that, but yeah. That you're right. The social... I don't know what you'd say. The social policy implications, the welfare implications, the, um, advertising role model, the, um, everything. The, just the downstream from this, you know, third order, fourth order, fifth order effects-

   19. SSSeth Stephens-Davidowitz28:16

       Yeah.

   20. CWChris Williamson28:16

       ... of us realizing this, uh, super, super profound. I hope ... You know, uh, w- we're only halfway through this, but one of the things that's certainly coming to mind at the moment for me is that you're right. What you've done so far, um, with your first book is kinda just laid out the facts, right? This is ... This, this is what it is. That's the surface level of the permafrost, right? But the roots underneath, that's the job of, of a lot of people, several orders-

   21. SSSeth Stephens-Davidowitz28:42

       Yeah.

   22. CWChris Williamson28:42

       ... of magnitude more people than a guy with, with a R code and a Google Doc. Like-

   23. SSSeth Stephens-Davidowitz28:48

       Yeah. Yeah.

   24. CWChris Williamson28:48

       ... uh, it's, it's, it's-

   25. SSSeth Stephens-Davidowitz28:49

       And there are, there are, there are lots of people in this area, and, you know, data science is just exploding in all kinds of ways, and I think a lot of people are, you know ... I think, uh, definitely millennials or people younger than millennials are also looking ... You know, it seems like the values are shifting a little bit where it's less about just making money. So I think initially everyone's kind of like, "Oh, data science. That's a lucrative field. I can get a job at-"

   26. CWChris Williamson29:13

       Mm-hmm.

   27. SSSeth Stephens-Davidowitz29:13

       "... you know, getting more people to click on ads, or, uh, work, get a job in finance." Which, which is tota- are totally fine jobs, uh, and, you know-

   28. CWChris Williamson29:19

       Lo and behold, they're swimming through Pornhub data five years after they finish their degree. (laughs)

   29. SSSeth Stephens-Davidowitz29:25

       Uh, yeah. Right. But I think, you know, a- a lot of people are reaching out to me that they're, you know, that they're kind of bored of studying. You know, they, they get, they're, they like data science but they're kind of bored of s- of getting people to click on ads, and they're, you know, they, they feel kind of unfulfilled and lacking purpose, and I think there are ways to use this data, uh, towards social good as well. Uh-

   30. CWChris Williamson29:42

       Got you.
10. 29:43 – 34:01

    Elections and prediction: what search behavior can reveal about voting

    1. CWChris Williamson29:43

       What, what's your ... So we're, we're in 2020. You've just crossed the threshold into the year of the presidential election in America. I mean, this, this must be like you staring down the barrel of the Super Bowl for your, your industry, or is it that you're just gonna have inundated with loads of different types of statistics, and then you can measure them before and look at what happens afterwards? Have you got some plans for this year with that?

    2. SSSeth Stephens-Davidowitz30:04

       Yeah. I, I, I'll play around. I play a- I play around every election. There are definitely insights in the internet. You know, I, I think Nate Silver of FiveThirtyEight does a really good job of making predictions based on kind of, uh, all the data, so it can be hard to, you know, tough to, to beat his predictions.

    3. CWChris Williamson30:20

       Mm-hmm.

    4. SSSeth Stephens-Davidowitz30:20

       Uh, you know, he put, he put so much thought and care into building-

    5. CWChris Williamson30:23

       Didn't he ... Di- did you retweet something of his about ... Was it the Democratic nominees? Was that, like, this week?

    6. SSSeth Stephens-Davidowitz30:29

       Oh, yeah, yeah, yeah, right. Yeah, yeah, yeah.... yeah, the iD- id. Well, he kind of, th- he just built a model of the Democratic nominee and I just thought that this wasn't based on data. This was actually just based on m- my intuition that he's maybe underestimating Michael Bloomberg's chances just because, uh, you know, he's... His whole, uh, kind of in the past models, the past elections where he's building, uh, the models from, there's never been a candidate who's willing to just spend (laughs) billions of dollars to try to get elected. So I think we don't really know how that's going to play out and that kind of, you know, m- makes me say he shouldn't... You know, I think in his model Bloomberg's at 1% or, or so- something like that. Uh, and I, I kind of think, you know, maybe 5% to 10% just because, uh, the amount of resources he's going to throw at this problem-

    7. CWChris Williamson31:12

       Got you.

    8. SSSeth Stephens-Davidowitz31:12

       ... uh, couldn't be-

    9. CWChris Williamson31:13

       What's the... So have you looked at much data? I don't know whether you can predict this far out, um, but have you looked at much data for moving forward into 2020?

    10. SSSeth Stephens-Davidowitz31:21

        I think it's ti- I think th- this time out, this, this far, this time out, th- th- this far out, it's tough. Uh, you need, you need to look closer. Uh, you know, I think so many people just aren't really thinking about their election or, you know, a lot of it, a lot of the general election will come down to who's going to actually turn out to vote.

    11. CWChris Williamson31:37

        Mm-hmm.

    12. SSSeth Stephens-Davidowitz31:37

        And that, I think we're going to need to wait, you know, a few weeks before the election where we'll really get clues, uh, on who's actually going to turn out to vote.

    13. CWChris Williamson31:45

        You hear the, the story that, um, people decide as they cross the threshold into the voting booth, right? Like, it's like the decision that they actually feel like they make is that, but it must be so interesting to, to compare because you have exit polls that don't... You know, in the last election that you guys had, the exit polls didn't marry up tremendously well with what actually happened. But were you... Did you have be- more of an insight about what was going to go on or retrospectively would you have had-

    14. SSSeth Stephens-Davidowitz32:13

        Well, I, I, I said Tr- I said Trump was going to win, but I don't know if that's because I'm a genius or a pessimist, because I'm just always predicting-

    15. CWChris Williamson32:19

        All right. Uh, let's go, let's go with genius. Let's go with genius.

    16. SSSeth Stephens-Davidowitz32:21

        Okay. (laughs) Yeah.

    17. CWChris Williamson32:21

        Let's go with genius.

    18. SSSeth Stephens-Davidowitz32:21

        I'm always predicting horrible things are going to happen, but-

    19. CWChris Williamson32:23

        Yeah.

    20. SSSeth Stephens-Davidowitz32:23

        ... uh, there, there are some clues. So one of the interesting things that correlates with voting outcome is the order you put candidates in searches. So a lot of people, they search Trump Clinton polls or Clinton Trump polls-

    21. CWChris Williamson32:35

        Mm-hmm.

    22. SSSeth Stephens-Davidowitz32:35

        ... or Trump Clinton debate, Clinton Trump election. You can see if you actually look historically the order, if people put Trump Clinton first, they're much more likely to go Trump. If they put Clinton Trump fir- Clinton Trump, Clinton first, they're more likely to go Clinton. And that's kind of interesting because it might almost be subconscious. And you could imagine someone, like they've been searching Trump Clinton election, Trump Clinton polls, Trump Clinton debate, and then they say... if you ask them, they're undecided and they go to the, they go to the, uh, polling place and they think they're undecided and then a few seconds before they cast their vote, as you said, they say, "Okay, you know, I'm feeling Trump." Well, maybe they weren't undecided all along and if you, like, looked at the data-

    23. CWChris Williamson33:15

        Mm-hmm.

    24. SSSeth Stephens-Davidowitz33:15

        ... they were giving away subconsciously, uh, which way they were going to go, um, from-

    25. CWChris Williamson33:20

        Have you, um, have you, have you looked at much of Sam Harris's work on free will?

    26. SSSeth Stephens-Davidowitz33:25

        Uh, I, I do know, I do know, I do know some of that work. Yeah.

    27. CWChris Williamson33:28

        Yeah. There's some interesting stuff on that about when they say, uh, raise your left or right hand or like pick a random city or whatever it might be, and if they put people into, uh, is it FMRIs where they can kind of do brain scans and stuff like that, and they're able to tell when someone made the decision to do the thing they're going to do before-

    28. SSSeth Stephens-Davidowitz33:46

        Right.

    29. CWChris Williamson33:46

        ... they do it and also before they think that they realized when they were going to decide to do it. Um, and that's kind of... This is like a more drawn out version of that, right?

    30. SSSeth Stephens-Davidowitz33:57

        Yeah. Definitely. I'm sure it happens. Uh, yeah, I'm sure it's a pretty widespread phenomenon.

11. 34:01 – 38:27

    Dating insights from recorded speed dates: what predicts a second date

    1. CWChris Williamson34:01

       Mm. Yeah. So there's... I'm looking at some of the things that I pulled out of the book. Can we, can we talk about, um, what, what should you say on a first date if you want a second?

    2. SSSeth Stephens-Davidowitz34:11

       Oh, yeah. (laughs) Well, that's not my study. That's a study of other researchers. They actually had people tape... They were in speed... They were people going through speed dates and they had people tape their dates, uh, like tape record everything that was said.

    3. CWChris Williamson34:23

       Mm.

    4. SSSeth Stephens-Davidowitz34:23

       And then after the date, it was heterosexual dates, the man and the woman said basically whether they wanted to go on a second date. Uh, so they could actually correlate the words said on the date with the probability that both you like the other person and that the other person likes you.

    5. CWChris Williamson34:38

       Mm.

    6. SSSeth Stephens-Davidowitz34:40

       So some of the things weren't that surprising. So for example, uh, if a woman laughs at a man's jokes, she's more likely to, to like him, to want to go on a second date with him

    7. NANarrator34:51

       Mm-hmm.

    8. SSSeth Stephens-Davidowitz34:51

       ... and this is pretty well, well known. Uh, men to get a woman to like them are... They're supposed to use words that show kind of support and care. So they say things like, "That must have been tough," or, uh, "Dat- that- that must have been tough," or, you know, "That sounds hard." Uh, that kind of increases the probability that a woman, uh, likes you. And then s- sometimes people like give away, again, subconsciously how they feel, or maybe not so subconsciously, but-

    9. CWChris Williamson35:20

       Mm-hmm.

    10. SSSeth Stephens-Davidowitz35:20

        ... if a woman uses what are called hedge words, so she says things like maybe or kinda, uh, then she's much less likely to like the guy.

    11. CWChris Williamson35:30

        Regardless of what-

    12. SSSeth Stephens-Davidowitz35:32

        Yeah, regardless of the content.

    13. CWChris Williamson35:32

        ... the topic is. Maybe I like cheese, maybe I want a dessert.

    14. SSSeth Stephens-Davidowitz35:36

        "Do you want a second date?" "Uh, kinda," whatever, it's kind of like she's kind of giving away that she's not excited about the guy in, in, in saying those things. So that's, that's pretty interesting and then, uh, the more a woman talks about herself, the more she likes the guy and the more likely the guy is to like her. Uh, so-

    15. CWChris Williamson35:55

        That's interesting.

    16. SSSeth Stephens-Davidowitz35:56

        ... like a good successful date tends to have more, more... The woman uses "I" more. Uh, uh, the, the more the woman uses "I", the more likely it's a successful date from both people's perspective. Uh, so kind of the conve- conversation shifts towards the woman's life. That's a good, a good sign.

    17. CWChris Williamson36:11

        Mm. I wonder if, uh, I wonder if Neil Strauss could rewrite the game or could do like the game 2.0, uh, but just do it off the back of like big data analysis? I mean-

    18. SSSeth Stephens-Davidowitz36:20

        I think it's def- it's definitely promising. Uh, you know, de- definitely a lot of these, uh, you know, I think, yeah, I think a lot of... a lot of the roles that people have come up with, uh-... you know, so- some of them are probably true. Some of them, you find out in the data aren't true, uh, so, yeah. I think-

    19. CWChris Williamson36:41

        It's i- it's interesting to have some of the sort of, I don't know whether you'd call it, like, folklore or some of the stuff that people posit about human nature, right? Like, you have from Neil Strauss, just a guy that tries to teach someone pick apart history with real world experience to a doctor who's got a degree in psychiatry, psychology or philosophy or, you know, human behavior, b- behavioral economics, whatever it might be. For the most part, people are just kind of creating these proxies or this, like, closest justifiable reason for why they think someone does something and trying to link these two things together. And then there's you, who's kind of just got this X-ray screen that actually gets to look at precisely what it is without them being worried about signaling to a researcher, without them being worried about it coming back to, uh, uh, haunt them at work because they said that ... because they, uh, agreed that they voted for some terrible, uh, political party, uh, with bad views or whatever it might be. It must be ... Uh, there'll be a lot of, um, fields, I think, that might end up becoming ousted or, like, you know, kind of really, really upended with some of the things that big data will come up with.

    20. SSSeth Stephens-Davidowitz37:52

        Well, it's not just me, but yeah, I definitely-

    21. CWChris Williamson37:54

        Uh, yeah, it's not only ... Partly you. (laughs)

    22. SSSeth Stephens-Davidowitz37:56

        ... I definitely agree that that, uh ... This type of research is very powerful tool. Uh, you know, sometimes it just confirms what people, you know, previously thought. Again, that, that a woman laughs at a man's joke, that, that, that means she likes him. I think that's one that we kind of figured out, uh, without data analysis, without tape recording-

    23. CWChris Williamson38:13

        Mm-hmm. Mm-hmm.

    24. SSSeth Stephens-Davidowitz38:13

        ... uh, everybody's ... All the, all these dates and, uh, fol- and, and, you know, uh, mining the text, but, uh, there definitely are areas where I think, you know, our intuition, uh, and our theories have been wrong 'cause they haven't been based on data.

12. 38:27 – 53:16

    The next book: data-driven life decisions (happiness, parenting, neighborhoods)

    1. CWChris Williamson38:27

       Mm-hmm. So what ... Y- you said that you're working on a second book at the moment. What's some of your research been on? What have you been interested in recently?

    2. SSSeth Stephens-Davidowitz38:34

       Uh, so my second book is on how you can use data to make better life decisions, kinda going off this idea of what you should say on a first date.

    3. CWChris Williamson38:42

       Mm-hmm. (laughs)

    4. SSSeth Stephens-Davidowitz38:42

       Uh, partly based on ... Part of the, the motivation for my book is that, uh, when I read the ... When I, when I, uh ... What, what ... You can actually see now, you can get kind of data on what people, what registers with people as they read your book because on Amazon Kindle, you can see the most underlined lines. So you can-

    5. CWChris Williamson39:02

       Oh, the highlights. Highlight function, yeah.

    6. SSSeth Stephens-Davidowitz39:04

       Highlight. Exactly. So one of the things I noticed is that people seem really interested in ways they can improve their own lives-

    7. CWChris Williamson39:09

       (laughs)

    8. SSSeth Stephens-Davidowitz39:10

       ... uh, which kind of fits the everybody lies theme because I think people don't necessarily like to admit that as much. People don't ... Particularly intellectuals, the type of people who would be drawn to my book, like to say they don't read for self-help, they read to learn more about the world or to, uh, kind of help other people, and I think you definitely do see people want to know, "What, what can I say on a first date? How can I make more money? Uh, you know, when should ... How, how can ... What, what, what business should I start?" Uh, so based on that, I'm basically just catering to the masses and writing a book on how you can-

    9. CWChris Williamson39:38

       Okay.

    10. SSSeth Stephens-Davidowitz39:38

        ... use data for better life decision (laughs) -

    11. CWChris Williamson39:40

        Yeah, you-

    12. SSSeth Stephens-Davidowitz39:40

        ... decisions.

    13. CWChris Williamson39:41

        Uh, you can use data to work out what you should write in a book to have another New York Times bestseller.

    14. SSSeth Stephens-Davidowitz39:47

        Yeah. (laughs) Yeah. I guess that's ... That, that ... Yeah, I guess that's the motivation. But so I'm going through all the different areas of life and parenting and dating and happiness and, uh ...

    15. CWChris Williamson39:56

        Wow. So have you, have you elicited any interesting insights into parenting or happiness recently?

    16. SSSeth Stephens-Davidowitz40:01

        Yeah, well, I mean, I think a lot of it's researching other people's data but ... Uh, uh, well, talking about other people's data that I don't think n- not enough people know about, but I'm really fascinated by these people who do happiness studies. I don't know if you've heard about this. They ask people, they ping people different times of the day and they ask them what they're doing and their mood. And-

    17. CWChris Williamson40:18

        Oh, yeah. Yeah, yeah. I have, I have heard about them. Yeah.

    18. SSSeth Stephens-Davidowitz40:20

        Yeah. So I think, uh, that's just kind of a really fascinating window into, uh, human happiness, uh, that I think hasn't been fully, you know, hasn't been fully discussed, so there are all these kind of interesting thing. And there are, like, some kind of interesting counterintuitive things. So for example, when, when people are drinking alcohol, they get a big boost in happiness. That's not surprising. Uh, so you get, like, three or four points on a hundred point scale of happiness if you're drinking-

    19. CWChris Williamson40:44

        Mm-hmm.

    20. SSSeth Stephens-Davidowitz40:45

        ... uh, al- alcohol. Uh, but one of the things that's interesting is that we tend- people tend to drink alcohol when they're doing something already fun, so if they're doi- if they're socializing with their friends or they're having sex, then they, uh, tend to drink alcohol, but actually if you drink alcohol then, it only gives you a tiny boost in happiness or no boost. But if you drink alcohol when you're doing something boring, which people, like, never do, or, or much less, less likely to do, then you get a huge boost in happiness. So you actually, like, uh, get a bigger ... If you're, like, cleaning, you know, wa- like, cleaning up, sweeping or the floor or something or commuting to work.

    21. CWChris Williamson41:19

        Mm-hmm.

    22. SSSeth Stephens-Davidowitz41:19

        Uh, you know, obviously there are reasons sometimes not to, not to drink during these opportunities, but, uh, it- it's actually more effective for your happiness. I think people use alcohol to try to take a good experience or great experience and make it epic, which doesn't really work, uh, instead of kind of, uh, kind of break- avoiding the doldrums, kind of low-

    23. CWChris Williamson41:40

        Yeah. It, it's kind, it's kind of unaesthetic, right?

    24. SSSeth Stephens-Davidowitz41:42

        ... working on fun. But if you ... Yeah, which is dangerous advice. It's kind of a path towards alcoholism, so I wanna be very careful that, like, a lot of people get addicted to alcohol and I don't want to just, you know, just telling people to drink any time they're bored or unhappy, uh, isn't nece- it's, you know, you have to use caution, but it is kind of ... It is, it does kind of show ... I think one of the things I'm trying to show is ways that we may, we may counterintu- like, the ways that counterintuitive decisions, kind of things that don't feel right can be right. So it doesn't feel right when, when you're having fun with friends, uh, you know, or you're doing something really fun, you really feel like, "If I ha- start drinking, this is gonna take it, this is gonna take me to the next level," and it tends not to work out that way. Uh, you know, so, so, but, you know, but, but, you know, you, you don't always think, "Oh, you know, if I have a beer or two now when I'm doing something, you know, just, just doing this really boring thing, then I, you know, then, then it'll be just fine and fun and I'll, I'll be all right," so, uh-

    25. CWChris Williamson42:35

        That's kind of ... That's really interesting. It's, it's so, so fascinating that-... it's a small proportion of the enjoyment from typical drinking activities come from the drinking, and most of it comes from the activities. The-

    26. SSSeth Stephens-Davidowitz42:48

        Yeah, exactly.

    27. CWChris Williamson42:48

        ... playing darts, the pool, the whatever.

    28. SSSeth Stephens-Davidowitz42:49

        That's, uh, that's probably (inaudible) . That, yeah, that's, that's probably one of the reasons we think drinking with friends is so much fun, because drinking with friends is so much fun. But being sober with friends also is so much fun (laughs) . So, uh, yeah. So we don't kind of distinguish that, you know, that a lot of it, the r- reason it's so fun is 'cause of the activity and we could just do it without it.

    29. CWChris Williamson43:07

        Oh, fascinating.

    30. SSSeth Stephens-Davidowitz43:07

        So like, for instance, yeah, like, like, uh, you get a huge bump in happiness if you drink while you're, like, getting ready to go out rather than when you actually go out. Like, 'cause that's boring.
13. 53:16 – 59:23

    Other datasets and controversies: abortions, racism, Wikipedia births, Facebook fandom

    1. CWChris Williamson53:16

       Yeah, for sure. Um, did you do some stuff on the stock market as h- having a, a little bit of a look? Did you look at, uh, the way that the stock market moves in gaming the stock market?

    2. SSSeth Stephens-Davidowitz53:25

       Y- I, I have a little bit. I, I, I didn't have t- too much success. It's pretty tough. Stock market's pretty, uh, pretty chaotic. And I think, uh, you know, one of the conclusions I come up with in my book is that it's a lot easier to find insights into like racism or child abuse or abortion or, uh, these other areas 'cause unfortunately there's not as much talent trying to find those insights as the stock market.

    3. CWChris Williamson53:49

       Yeah. (laughs)

    4. SSSeth Stephens-Davidowitz53:50

       So the stock market, you're competing against, you know, astrophysicists and phys- e- e- everybody's trying to figure out the stock market. And, you know, so it's, it's a little bit of a, you know, a little bit more challenge. But, uh, I think-

    5. CWChris Williamson54:02

       Yeah, you can't trade, you can't trade racism on the open market, can you? So ...

    6. SSSeth Stephens-Davidowitz54:06

       Yeah, yeah. I've-

    7. CWChris Williamson54:06

       Yeah.

    8. SSSeth Stephens-Davidowitz54:06

       And, uh, yeah, you can't get rich, uh, finding a new map of racism in the United States. Uh, you can get rich finding a hidden inefficiency in the stock market. So, uh, I think there's definitely much more talent, uh, towards the stock market that it, it can, it can be tough with public data-

    9. CWChris Williamson54:22

       Mm.

    10. SSSeth Stephens-Davidowitz54:23

        ... as I've historically used, uh, finding kind of an insight.

    11. CWChris Williamson54:26

        Mm-hmm. What about, you, you mentioned, uh, abortions there. There was the, the thing about a, the back alley, uh, uh, sort of abortion-

    12. SSSeth Stephens-Davidowitz54:34

        Yeah.

    13. CWChris Williamson54:34

        ... crisis thing. Can you take us through that?

    14. SSSeth Stephens-Davidowitz54:35

        Yeah, well I was just shocked by how frequently people search on Google for do it yourself abortions, for kind of, uh, giving yourself a miscarriage, giving yourself an abortion.... uh, and these searches are almost, uh, highly concentrated, almost perfectly ... If you look at where these searches are, almost perfectly mapped, a place where it's hard to get an abortion. And they got ... They went up a lot in 2011 when it ... there was kind of a crack down in the United States against legal abortion. So, I think ... And also if you actually look at the data, it does seem like there are missing pregnancies in those areas. So, uh, births, uh, basically births have gone down a lot and abortions haven't kind of ... Uh, abortio- abortions kind of also gone down a- and you kind of do the math, it seems like probably, uh, there, there are, are s- there is kind of a missing-

    15. CWChris Williamson55:20

        Something's happening somewhere.

    16. SSSeth Stephens-Davidowitz55:22

        ... happening, and I think a lot of that probably is, uh, off the books abortion. Some of it's not bad guy ... There are, there are sadly people who literally search how to use a coat hanger to give yourself an abortion, uh, but some of it's kind of abortion pills, uh, which now people are getting online, uh, which it ... Some people say it's actually a good thing. Uh, some abortion rights activists say it's actually a good thing, because they're pretty safe, uh, and they're a way for people, uh, who don't ... who are in areas where it's hard to get a legal abortion to kind of still have an abortion without other pe-

    17. CWChris Williamson55:54

        Without birth control, yeah, yeah, yeah. Ah.

    18. SSSeth Stephens-Davidowitz55:57

        It's not, it's not, uh, a huge disaster unnecessarily, uh, but ...

    19. CWChris Williamson56:01

        Got you. So, most of what we've spoken about today so far has been to do with Google. Were there any other platforms that you were able to pull data from, that you could look at?

    20. SSSeth Stephens-Davidowitz56:10

        Yeah. So Pornhub, as I talked about. I analyzed Stormfront, that's kind of a white supremacist site. Uh, also an interesting data set. Uh, that's it, uh, I analyzed Wikipedia, where successful people tend to be born. Uh, Facebook.

    21. CWChris Williamson56:25

        Where do they ... Where are they born?

    22. SSSeth Stephens-Davidowitz56:27

        College towns and cities, uh, mostly. So, uh, and yeah, that's, that's, um, yeah. (laughs) That, that, and part of that's kind of the genetics. They probably have ... You know, if you're a kid of, uh, professors you're more likely to be smart yourself and, uh, be more likely to be notable in various ways. But I think some of it is exposure to innovation. Again, expo- you know, early exposure. So you see rock and roll stars are much more likely to grow up in college towns, and I think the part, part of the reason for that is because college towns are kind of places of musical innovation. Uh, the rec- the, uh ... They have, you know, they, they historically ... It's kind of changing but they had these kind of, uh, uh, they had record stores that were kind of cutting edge, cut ... People would, you know, really cool bands would come and play and, uh, uh, they had good, uh, radio stations.

    23. CWChris Williamson57:22

        Mm-hmm.

    24. SSSeth Stephens-Davidowitz57:22

        So I think part of it, the early exposure to innovation.

    25. CWChris Williamson57:25

        Got you. And what about Facebook? We got ... Got anything cool that you've realized off Facebook recently that you've had data from?

    26. SSSeth Stephens-Davidowitz57:33

        One of the things I did is a little politically incorrect, but I looked at ... So Facebook you can measure. For basketball I looked at how many fans everybody has on Facebook. Uh, so I looked for basketball players basically. How many fans basketball players have.

    27. CWChris Williamson57:46

        Mm-hmm.

    28. SSSeth Stephens-Davidowitz57:47

        I wanted to see if, h- how white fan- white basketball players compared to black basketball players. So historically, it's been thought that kind of white players get kind of a boost in fandom and that a lot of teams have thought to ... We thought that a lot of teams hired like as their 12th player, their, their, their lowest bench player, a white guy, just because-

    29. CWChris Williamson58:07

        The token white guy.

    30. SSSeth Stephens-Davidowitz58:09

        Yeah, a token white guy because people would be more likely to go ... So you can actually c- You can imagine building an analysis where you, you know basically how good every player is in basketball. You can control for all their stats. So how many points they score, how many rebounds they have, how many assists they have, et cetera. And you say like, controlling for that, all else equal, how many fans do they have on Facebook? And what you see is that African American players just have w- way more fans, mostly due to a huge bump up among African Amer- They basically get a little bump from everybody, so they're a little more popular among white people.
14. 59:23 – 1:02:01

    Wrap-up: anonymity, personal tech habits, and where to find Seth

    1. CWChris Williamson59:23

       Got you. So the final question that I wanted to ask is actually one from, uh, Jordan, who's part of the Modern Wisdom project. And he was saying, have your insights from big data changed your use of technology or ...

    2. SSSeth Stephens-Davidowitz59:37

       I think the only thing is I Google myself more now that I'm in-

    3. CWChris Williamson59:40

       (laughs) What an answer.

    4. SSSeth Stephens-Davidowitz59:44

       Yeah. In news pa- but otherwise I think, I don't really think that it's made a big change.

    5. CWChris Williamson59:48

       Got you, yeah. Oh, it's ... (clears throat) One of the things certainly that's come up here is, it is a, a very, um, interesting insight into human nature. Into what it is that we do and all the stuff like that. But as you say, this anonymous aggregated data is precisely that. Like, it's happening, but I don't know that it's you or as if it's the guy next door.

    6. SSSeth Stephens-Davidowitz1:00:11

       Yep.

    7. CWChris Williamson1:00:12

       Yeah.

    8. SSSeth Stephens-Davidowitz1:00:12

       Yeah. Definitely. (laughs)

    9. CWChris Williamson1:00:14

       Yeah, which is interesting. Anyway, man, this has been absolutely awesome. Have we got ... Seth, have we got a, um, a, a date or a, uh, an idea in mind about when the, the next book's gonna come out?

    10. SSSeth Stephens-Davidowitz1:00:24

        Probably 2021, but uh, I'm not sure exactly when.

    11. CWChris Williamson1:00:28

        You got to get through this presidential election, then you're gonna have to do all of your cool-

    12. SSSeth Stephens-Davidowitz1:00:32

        Yeah, I think, uh, yeah. I'll publish it during the election because everyone's gonna be focused on the, on the election.

    13. CWChris Williamson1:00:37

        Yeah. I was talking to, uh, Paul Bloom, uh, from Yale and, uh-

    14. SSSeth Stephens-Davidowitz1:00:41

        Okay.

    15. CWChris Williamson1:00:41

        He was, he was saying precisely the same thing. He's doing this new book about, um, suffering. About how people really enjoy suffering. He'd managed to find a link between BDSM and meditation, which actually sounds exactly like one of the things that you would, you would've come up with out of big data. And, um, he wa- he was saying ex- He was like, "Ah, I think it'll maybe be finished, like, first draft sort of start of the year," and blah, blah, blah. And I said, "Oh cool. So are we gonna get it next year?" And he's like, "No, man. (laughs) It's an election this year, like (laughs) I'm not, I'm not releasing anything."

    16. SSSeth Stephens-Davidowitz1:01:09

        Right.

    17. CWChris Williamson1:01:10

        (inhales through nose) So 2021, hopefully we'll get a load of new, uh, literacy stuff through. So, um, where can people find you online, Seth? They want to f- follow your stuff. Where can they go?

    18. SSSeth Stephens-Davidowitz1:01:21

        Uh, probably just Google Seth Everybody Lies. Nobody's going to remember my last name, so just Seth Everybody Lies and then they'll, they'll find, you know, my Twitter and everything else. So ...

    19. CWChris Williamson1:01:31

        Awesome, man.

    20. SSSeth Stephens-Davidowitz1:01:31

        Yeah.

    21. CWChris Williamson1:01:31

        Thank you so much. Well, I really appreciate your time, um, um, I'm gonna try and work out who it was that I lived near at home and, uh, and see where the, the in- influence was on me.

    22. SSSeth Stephens-Davidowitz1:01:44

        Great.

    23. CWChris Williamson1:01:44

        Cheers, man. Thank you so much, dude.

    24. SSSeth Stephens-Davidowitz1:01:48

        Thank you.

    25. NANarrator1:01:50

        (instrumental music plays)

Episode duration: 1:02:01