1 00:00:05,640 --> 00:00:08,640 Speaker 1: Hi Brian, Hi Katie Brian. The top website in the 2 00:00:08,680 --> 00:00:10,840 Speaker 1: US is Google. Number two is YouTube. Do you know 3 00:00:10,880 --> 00:00:13,119 Speaker 1: what number three is? Well, if you'd asked me up 4 00:00:13,160 --> 00:00:16,120 Speaker 1: until recently, I would have said maybe Facebook or Amazon. 5 00:00:17,079 --> 00:00:19,200 Speaker 1: You know, you would have been wrong, because it turns out, 6 00:00:19,239 --> 00:00:22,760 Speaker 1: according to the most recent data, the number three website 7 00:00:22,760 --> 00:00:25,880 Speaker 1: in America is actually read It, which I think we 8 00:00:25,920 --> 00:00:28,600 Speaker 1: were both pretty surprised to hear. Yeah, and that brings 9 00:00:28,680 --> 00:00:31,000 Speaker 1: us to our guest today, Reddit CEO and one of 10 00:00:31,040 --> 00:00:34,559 Speaker 1: its co founders, Steve Huffman. We taped this podcast in 11 00:00:34,600 --> 00:00:38,440 Speaker 1: front of a live audience at the wonderful Aspen Ideas Festival. 12 00:00:38,640 --> 00:00:41,519 Speaker 1: I love this festival. I've come every year, I think 13 00:00:41,520 --> 00:00:44,919 Speaker 1: for the past decades, so clearly it's one of my favorites. 14 00:00:45,120 --> 00:00:47,400 Speaker 1: Now many, if not most, of the people who use 15 00:00:47,479 --> 00:00:50,720 Speaker 1: Reddit are in their teens and twenties. So for my 16 00:00:50,840 --> 00:00:53,520 Speaker 1: peers and others who may not be that familiar with it, 17 00:00:53,800 --> 00:00:55,720 Speaker 1: let me get you up to speed. Reddit is a 18 00:00:55,760 --> 00:00:58,880 Speaker 1: site that's made up of lots of different communities called 19 00:00:59,080 --> 00:01:03,120 Speaker 1: sub reddits, where people can post articles, videos, thoughts. The 20 00:01:03,160 --> 00:01:05,399 Speaker 1: sky's the limit really, and there are actually over a 21 00:01:05,480 --> 00:01:09,280 Speaker 1: million subreddits devoted to everything from world news to crafts 22 00:01:09,360 --> 00:01:12,120 Speaker 1: to life hacks like how do you get abs? Like 23 00:01:12,160 --> 00:01:17,680 Speaker 1: Brian Goldsmith answers it's very hard, very funny. The site 24 00:01:17,720 --> 00:01:22,240 Speaker 1: actually has remarkably simple interface, almost dated really, And I 25 00:01:22,319 --> 00:01:25,200 Speaker 1: was very excited to talk to Steve Reddit CEO because 26 00:01:25,200 --> 00:01:28,399 Speaker 1: the site has found itself in the crosshairs of some 27 00:01:28,680 --> 00:01:31,880 Speaker 1: very thorny issues these days. So we spoke with Steve 28 00:01:31,920 --> 00:01:34,800 Speaker 1: about everything from fake news to Russian bods, to pizza 29 00:01:34,800 --> 00:01:37,640 Speaker 1: gate to hate speech, and we started by asking how 30 00:01:37,680 --> 00:01:43,800 Speaker 1: would he describe Reddit? Even at Reddit, this is something 31 00:01:43,880 --> 00:01:46,520 Speaker 1: we debate a lot of what redd it is and 32 00:01:46,560 --> 00:01:49,520 Speaker 1: how to explain Reddit to folks, but effectively read it 33 00:01:49,560 --> 00:01:51,880 Speaker 1: as a network of communities. And so we have communities 34 00:01:51,920 --> 00:01:56,920 Speaker 1: that cover pretty much any topic imaginable. So things that 35 00:01:56,920 --> 00:02:02,320 Speaker 1: that are very recognizable, so news, sports, politics, um, pretty 36 00:02:02,400 --> 00:02:07,760 Speaker 1: much any interest and passion. And birds with arms is 37 00:02:08,680 --> 00:02:11,320 Speaker 1: would be up there. Um birds with arms as a 38 00:02:11,320 --> 00:02:16,560 Speaker 1: community where one photoshops arms on the birds. Um. No, 39 00:02:16,680 --> 00:02:20,080 Speaker 1: it's this is important stuff. I'm glad you're here. Um. 40 00:02:21,919 --> 00:02:25,240 Speaker 1: And then we also have communities. Um uh, excuse me, 41 00:02:25,320 --> 00:02:27,640 Speaker 1: I cleaned myself this morning. I have a hairball. Um. 42 00:02:29,440 --> 00:02:35,240 Speaker 1: Communities uh dedicated to UH support in helping each other out. 43 00:02:35,280 --> 00:02:38,080 Speaker 1: Some communities for new parents who are maybe struggling with that, 44 00:02:38,120 --> 00:02:40,800 Speaker 1: our communities for people are struggling with addiction or other 45 00:02:40,840 --> 00:02:44,400 Speaker 1: difficult times in their life. And so really any topic 46 00:02:44,480 --> 00:02:48,040 Speaker 1: around UH, you know, the things that humans will gather 47 00:02:48,120 --> 00:02:50,960 Speaker 1: around a communicate around, is represented on Reddit somewhere. And 48 00:02:51,000 --> 00:02:55,160 Speaker 1: so our mission Reddit is to bring UH community in 49 00:02:55,200 --> 00:02:57,560 Speaker 1: a sense of belonging to everybody on the planet. So 50 00:02:57,600 --> 00:03:00,600 Speaker 1: since people in this room no offense to you, all 51 00:03:01,040 --> 00:03:04,919 Speaker 1: may not be read it's demographic. Who is Rennet's demographic? 52 00:03:05,160 --> 00:03:08,520 Speaker 1: And why do you think it caught fire so quickly 53 00:03:08,680 --> 00:03:12,800 Speaker 1: at the outset? Sure, so uh you all are are 54 00:03:12,880 --> 00:03:18,880 Speaker 1: maybe not our primary demographic, but um we read it 55 00:03:19,000 --> 00:03:21,640 Speaker 1: is broad enough now that that pretty much everybody has 56 00:03:21,680 --> 00:03:24,640 Speaker 1: represented on Reddit somewhere. So I guarantee all of you 57 00:03:24,680 --> 00:03:27,600 Speaker 1: have a home on Reddit somewhere the way Redd has 58 00:03:27,639 --> 00:03:30,720 Speaker 1: grown our origin. So we're thirteen years old. We started 59 00:03:31,000 --> 00:03:33,040 Speaker 1: in two thousand five. My college roommate and I started 60 00:03:33,080 --> 00:03:36,400 Speaker 1: read it immediately after we graduated, and at the time 61 00:03:36,400 --> 00:03:40,320 Speaker 1: read it was just a pile of links, no categorization, 62 00:03:40,800 --> 00:03:44,760 Speaker 1: no communities, no no commenting even and what what made 63 00:03:44,800 --> 00:03:47,680 Speaker 1: redd at work. Then it was very, very serendipitous. And 64 00:03:47,720 --> 00:03:50,880 Speaker 1: the reason we grew is because both those that serendipity, 65 00:03:50,880 --> 00:03:52,640 Speaker 1: you never knew what you're gonna find. You're gonna find 66 00:03:52,800 --> 00:03:55,040 Speaker 1: some politics, Are you're gonna find an Internet meme? Are 67 00:03:55,040 --> 00:03:58,680 Speaker 1: you going to find commentary critical of Reddit or something else? 68 00:03:59,360 --> 00:04:03,560 Speaker 1: And it was very genuine and authentic, and we resonated 69 00:04:03,640 --> 00:04:06,120 Speaker 1: with I think I think a generation of people who 70 00:04:06,680 --> 00:04:10,560 Speaker 1: had just become kind of skeptical and jaded of of everything, right, 71 00:04:10,600 --> 00:04:14,080 Speaker 1: of all media, of news, of Anything's just this feeling 72 00:04:14,120 --> 00:04:16,320 Speaker 1: of like everything is bullshit, I'm always being lied to, 73 00:04:16,920 --> 00:04:19,200 Speaker 1: And so we still have that aspect to read it, 74 00:04:19,240 --> 00:04:22,440 Speaker 1: and so we still um pick up users in high 75 00:04:22,480 --> 00:04:24,960 Speaker 1: school and college. I think when when when young adults 76 00:04:24,960 --> 00:04:27,560 Speaker 1: are turning into adults and they're starting to develop kind 77 00:04:27,600 --> 00:04:31,360 Speaker 1: of worldviews and skepticism and critical thinking, read it is 78 00:04:31,360 --> 00:04:33,680 Speaker 1: a great place for them. But then we tend to 79 00:04:33,760 --> 00:04:36,080 Speaker 1: keep them, you know, and and and so we're thirteen 80 00:04:36,120 --> 00:04:38,479 Speaker 1: years old now, so I would say our our media 81 00:04:38,560 --> 00:04:42,359 Speaker 1: age is probably you know, let's call it three, but 82 00:04:42,400 --> 00:04:44,839 Speaker 1: our average age is increasing over time. So people who 83 00:04:44,839 --> 00:04:47,120 Speaker 1: want kind of an outlet to express their views and 84 00:04:47,160 --> 00:04:50,760 Speaker 1: to share it with other people are attracted to read it. 85 00:04:50,839 --> 00:04:54,039 Speaker 1: But like all of their social platforms, you all have 86 00:04:54,279 --> 00:04:57,480 Speaker 1: been going through your fair share of controversy, and we're 87 00:04:57,520 --> 00:05:00,080 Speaker 1: gonna talk about some of the challenges that you all 88 00:05:00,080 --> 00:05:02,040 Speaker 1: had faced as a platform in a moment. But first 89 00:05:02,080 --> 00:05:04,640 Speaker 1: we want to talk about how this whole thing began. 90 00:05:05,080 --> 00:05:08,240 Speaker 1: You started Reddit. You and your college roommate founded the 91 00:05:08,240 --> 00:05:11,719 Speaker 1: company just a few weeks after graduating from the University 92 00:05:11,720 --> 00:05:14,400 Speaker 1: of Virginia back in two thousand five. So how did 93 00:05:14,400 --> 00:05:18,640 Speaker 1: you come up with the idea? Well, almost accidentally, my 94 00:05:18,720 --> 00:05:22,360 Speaker 1: roommate Alexis Hanion and I we we have been roommates 95 00:05:22,400 --> 00:05:24,760 Speaker 1: and housemates all through college, and on our last year 96 00:05:24,760 --> 00:05:27,520 Speaker 1: of college, we started talking about companies that we might 97 00:05:27,560 --> 00:05:32,360 Speaker 1: start together. And at the same time an idol of mine, 98 00:05:32,400 --> 00:05:35,840 Speaker 1: this fellow Paul Graham, who he started a company of 99 00:05:35,839 --> 00:05:37,799 Speaker 1: his own that sold during the bubble, and he would 100 00:05:37,800 --> 00:05:43,200 Speaker 1: write online evangelizing startups and and his whole journey, and 101 00:05:43,200 --> 00:05:46,120 Speaker 1: he eventually founded why comminat So he started by Combinator, 102 00:05:46,400 --> 00:05:49,080 Speaker 1: which at the time was a kind of a rag 103 00:05:49,160 --> 00:05:53,719 Speaker 1: tag group of investors and is now the largest siege 104 00:05:53,720 --> 00:05:55,640 Speaker 1: stage investment firm in the world. They do about four 105 00:05:56,000 --> 00:05:59,560 Speaker 1: companies a year. Back in two thousand five, we applied 106 00:05:59,560 --> 00:06:02,480 Speaker 1: to y C with a completely different idea. I had 107 00:06:02,520 --> 00:06:05,119 Speaker 1: to do with ordering food from from your cell phone. 108 00:06:05,440 --> 00:06:10,800 Speaker 1: We were wisely rejected this first one. Yes, yeah, it 109 00:06:10,880 --> 00:06:16,160 Speaker 1: was an idea actually ahead of its time. Um uh. 110 00:06:16,640 --> 00:06:19,920 Speaker 1: But but Paul and his partner Jessica basically reached out 111 00:06:19,960 --> 00:06:22,520 Speaker 1: to us and said, we really like you too. If 112 00:06:22,600 --> 00:06:24,960 Speaker 1: you want to be in this program, let's find something 113 00:06:25,000 --> 00:06:26,920 Speaker 1: else for you to do. So we had a brainstorming 114 00:06:26,960 --> 00:06:30,000 Speaker 1: session and in that session we basically decided we were 115 00:06:30,000 --> 00:06:32,160 Speaker 1: going to build I have this on my first page 116 00:06:32,160 --> 00:06:33,560 Speaker 1: and notes, We're going to build a front page of 117 00:06:33,560 --> 00:06:35,880 Speaker 1: the Internet. And that's something we used to say tongue 118 00:06:35,880 --> 00:06:39,240 Speaker 1: in cheek back then um, and I always say it um, 119 00:06:39,320 --> 00:06:42,320 Speaker 1: I think somewhat somewhat proudly now. But that was really 120 00:06:42,320 --> 00:06:44,960 Speaker 1: the genesis of the idea and our motivation in those 121 00:06:45,200 --> 00:06:48,320 Speaker 1: those first months was I didn't want to look stupid 122 00:06:48,320 --> 00:06:50,440 Speaker 1: in front of Paul Graham, and I didn't want my 123 00:06:50,560 --> 00:06:53,600 Speaker 1: mom to be right because she wanted me to get 124 00:06:53,600 --> 00:06:56,479 Speaker 1: a real job. And I was going up to Boston 125 00:06:56,520 --> 00:06:58,480 Speaker 1: to work with this weirdo Paul on this on this 126 00:06:58,560 --> 00:07:00,840 Speaker 1: Reddit thing. Fun fact, by the way, we like to 127 00:07:00,880 --> 00:07:03,800 Speaker 1: sprinkle our podcast with these. Your co founder, Alexis you 128 00:07:03,880 --> 00:07:07,520 Speaker 1: just mentioned, is now married to Serena Williams. That's pretty cool. 129 00:07:07,600 --> 00:07:11,160 Speaker 1: That is a fun fact. Do you and your wife 130 00:07:11,200 --> 00:07:13,720 Speaker 1: ever double date with them? I am no longer married. 131 00:07:14,520 --> 00:07:17,040 Speaker 1: Well that do you ever go out as a threesome? 132 00:07:19,600 --> 00:07:23,239 Speaker 1: Let's not misinterpret that question. Okay, do you ever tagal 133 00:07:23,360 --> 00:07:28,360 Speaker 1: on ever the third Wheel? Uh? No, very very seldomly 134 00:07:28,720 --> 00:07:31,640 Speaker 1: Alexis and Serna, I have honestly no idea where in 135 00:07:31,640 --> 00:07:33,840 Speaker 1: the world they are at any point in time. But 136 00:07:33,920 --> 00:07:36,040 Speaker 1: you're still good friends with Alexis. We've been friends a 137 00:07:36,080 --> 00:07:39,360 Speaker 1: long time. Yeah. Well, let's talk about after you launched 138 00:07:39,360 --> 00:07:41,720 Speaker 1: the company, because it was seventeen months later and you 139 00:07:41,800 --> 00:07:45,360 Speaker 1: sold it to Conde Nast of all places. You have 140 00:07:45,440 --> 00:07:48,360 Speaker 1: said that you regretted that decision. Why, well, I have 141 00:07:48,480 --> 00:07:50,920 Speaker 1: said that, and I've also backtracked off of that. I've 142 00:07:51,520 --> 00:07:53,600 Speaker 1: in fact changed my story on this a lot over 143 00:07:53,640 --> 00:07:59,800 Speaker 1: the years, you know, as it serves me, and really 144 00:07:59,800 --> 00:08:05,360 Speaker 1: in gendering trust there, mostly with perspective. So the context 145 00:08:05,400 --> 00:08:07,680 Speaker 1: in which we sold the company, this was in two 146 00:08:07,680 --> 00:08:10,640 Speaker 1: thousand six. As I mentioned, when we started the company, 147 00:08:10,680 --> 00:08:14,160 Speaker 1: we didn't have a vision or a mission. We were 148 00:08:14,200 --> 00:08:16,160 Speaker 1: a little dysfunctional. We didn't know where we were going. 149 00:08:16,200 --> 00:08:20,320 Speaker 1: The economy wasn't looking pretty on the horizon. And along 150 00:08:20,400 --> 00:08:22,680 Speaker 1: came Conde and asked offering to give us a bunch 151 00:08:22,680 --> 00:08:24,720 Speaker 1: of money for this thing, and so we said, well, 152 00:08:24,720 --> 00:08:27,320 Speaker 1: this seems like a pretty good idea, and and it's funny. 153 00:08:27,440 --> 00:08:31,440 Speaker 1: I went into that acquisition with the mentality of I 154 00:08:31,440 --> 00:08:35,240 Speaker 1: thought I was going to have the stereotypical Silicon Valley 155 00:08:35,280 --> 00:08:39,000 Speaker 1: startup journey, which is founder starts, company, founder sales, company, 156 00:08:39,280 --> 00:08:42,560 Speaker 1: founder leaves two weeks later. I end up staying for 157 00:08:42,600 --> 00:08:45,520 Speaker 1: another three years. I really really enjoyed working on Reddit, 158 00:08:45,800 --> 00:08:50,640 Speaker 1: and I left in two thousand nine because for a 159 00:08:50,679 --> 00:08:54,400 Speaker 1: couple of reasons. One, I felt stagnant um. I felt 160 00:08:54,400 --> 00:08:56,199 Speaker 1: like as a person I wasn't learning and growing, and 161 00:08:56,240 --> 00:08:58,640 Speaker 1: I felt like I was too young to have that 162 00:08:58,679 --> 00:09:00,600 Speaker 1: feeling at work. And it's any when I look back 163 00:09:00,600 --> 00:09:03,480 Speaker 1: in that time, I think that's like just absolutely insane, right, 164 00:09:03,480 --> 00:09:05,320 Speaker 1: because now I just feel like I have so much 165 00:09:05,320 --> 00:09:07,280 Speaker 1: to learn and I'm like so far behind. But at 166 00:09:07,280 --> 00:09:10,640 Speaker 1: the time that's how I felt. I also, uh, I 167 00:09:10,679 --> 00:09:13,880 Speaker 1: liked perspective. I didn't realize I didn't appreciate how special 168 00:09:13,880 --> 00:09:15,240 Speaker 1: red it was. It was the only thing we had 169 00:09:15,240 --> 00:09:18,160 Speaker 1: ever done, and and and so I had this perspective 170 00:09:18,160 --> 00:09:21,880 Speaker 1: that if you put something online, it'll just grow and 171 00:09:21,880 --> 00:09:23,199 Speaker 1: and and that was my experience with thout it. So 172 00:09:23,240 --> 00:09:24,560 Speaker 1: I was like, well, I want to do something else 173 00:09:24,600 --> 00:09:26,960 Speaker 1: that just grows, but this time actually makes money. So 174 00:09:27,040 --> 00:09:29,760 Speaker 1: I started a company with another buddy called hip Monk, 175 00:09:29,840 --> 00:09:33,120 Speaker 1: which did travel search and hip Monk was interesting because 176 00:09:33,120 --> 00:09:41,360 Speaker 1: we were not burdened with growth um and and so 177 00:09:41,440 --> 00:09:45,440 Speaker 1: that was a really good set of lessons I learned 178 00:09:45,440 --> 00:09:47,079 Speaker 1: there in that time, and actually as as as a 179 00:09:47,080 --> 00:09:48,840 Speaker 1: as a business person, learned a lot of really important 180 00:09:48,880 --> 00:09:51,520 Speaker 1: lessons about how to be discipline, about customers and product 181 00:09:51,559 --> 00:09:55,000 Speaker 1: development and growing a team and shrinking a team at 182 00:09:55,000 --> 00:09:56,600 Speaker 1: the same time. Right, it continued to grow and grow 183 00:09:56,640 --> 00:09:59,040 Speaker 1: and grow, and during this period when you were off 184 00:09:59,080 --> 00:10:02,320 Speaker 1: working on hip Monk, credit was still growing. It hired 185 00:10:02,559 --> 00:10:05,600 Speaker 1: Ellen Powell to be the CEO, and she was famous, 186 00:10:05,640 --> 00:10:10,439 Speaker 1: of course pursuing Kleiner Perkins, the venture firm for gender discrimination, 187 00:10:10,880 --> 00:10:12,599 Speaker 1: and when Pal got to read it, she took a 188 00:10:12,679 --> 00:10:16,760 Speaker 1: number of tough, tough steps. She banned revenge porn. For example, 189 00:10:17,240 --> 00:10:20,680 Speaker 1: she banned five sub credits that engaged in various forms 190 00:10:20,679 --> 00:10:24,760 Speaker 1: of harassment. About two users signed a petition asking her 191 00:10:24,800 --> 00:10:29,559 Speaker 1: to leave, which ultimately she did with the benefit of hindsight. 192 00:10:30,160 --> 00:10:32,840 Speaker 1: Do you think that she was treated unfairly at Reddit? 193 00:10:33,240 --> 00:10:37,440 Speaker 1: Do I? I? You know, I think Ellen had a 194 00:10:37,520 --> 00:10:40,040 Speaker 1: rough go at a readit um, some within her control 195 00:10:40,080 --> 00:10:43,680 Speaker 1: and some knots. And but I think that the general 196 00:10:44,640 --> 00:10:47,880 Speaker 1: criticisms of of of Ellen, both from the Reddit community 197 00:10:47,920 --> 00:10:51,680 Speaker 1: and from the kind of just greater community, we're really unfair. UM. 198 00:10:51,720 --> 00:10:54,200 Speaker 1: Having been on the receiving end of that myself as well, 199 00:10:54,280 --> 00:10:56,920 Speaker 1: I I know what that feels like. Um. Do you 200 00:10:56,960 --> 00:11:00,280 Speaker 1: think the criticisms were motivated by gender? I think there 201 00:11:00,320 --> 00:11:02,400 Speaker 1: was that dimension to it, And certainly I think when 202 00:11:02,440 --> 00:11:04,319 Speaker 1: you also look at it through the lens of her 203 00:11:05,160 --> 00:11:08,760 Speaker 1: very public lawsuit against Kleiner, that that only served to 204 00:11:08,800 --> 00:11:11,160 Speaker 1: amplify that that narrative, and I think that was really 205 00:11:11,200 --> 00:11:13,640 Speaker 1: unfortunate in light of a me too movement. Do you 206 00:11:13,679 --> 00:11:16,880 Speaker 1: look at what happened to her any differently? Have you 207 00:11:16,920 --> 00:11:19,400 Speaker 1: had an epiphany or did you always feel that this 208 00:11:19,600 --> 00:11:22,200 Speaker 1: was unfair. No. I always felt that way, you know, 209 00:11:22,320 --> 00:11:25,439 Speaker 1: and study because when when she was going through that, 210 00:11:26,040 --> 00:11:28,840 Speaker 1: I was not at Reddit. I was I was a civilian. 211 00:11:28,880 --> 00:11:32,520 Speaker 1: I was a user watching this, and it was very 212 00:11:32,559 --> 00:11:34,920 Speaker 1: frustrating to watch from a couple of dimensions. Right one 213 00:11:35,080 --> 00:11:37,640 Speaker 1: as somebody who knows Reddit really well. I was like, 214 00:11:38,080 --> 00:11:40,560 Speaker 1: oh my gosh, there are things you just need to 215 00:11:40,640 --> 00:11:42,520 Speaker 1: do that I think you can get ahead of this 216 00:11:43,120 --> 00:11:45,640 Speaker 1: because one of the things we we saw is or 217 00:11:45,679 --> 00:11:48,480 Speaker 1: always I observed from the outside, and I think Ellen 218 00:11:48,600 --> 00:11:51,520 Speaker 1: was starting to get there. It's the the volume of 219 00:11:51,559 --> 00:11:55,199 Speaker 1: the the vitriol was not in proportion to the number 220 00:11:55,200 --> 00:11:58,360 Speaker 1: of people behind it, and that, to me is a 221 00:11:58,360 --> 00:12:00,839 Speaker 1: technology problem, and that's something I read it. We've always 222 00:12:00,840 --> 00:12:02,120 Speaker 1: thought a lot about, and this is something that I've 223 00:12:02,120 --> 00:12:04,280 Speaker 1: always thought a lot about. So watching that was particularly 224 00:12:04,280 --> 00:12:06,920 Speaker 1: frustrating cause I was like, there are some solutions here. 225 00:12:07,559 --> 00:12:10,720 Speaker 1: At the same time, it was the that whole conversation 226 00:12:10,800 --> 00:12:13,800 Speaker 1: was just I think, very unproductive and toxic and and 227 00:12:13,880 --> 00:12:15,360 Speaker 1: I don't I don't know if we really got anywhere 228 00:12:15,360 --> 00:12:17,640 Speaker 1: with it, to be honest, either, Silicon Valley is you know, 229 00:12:17,800 --> 00:12:21,400 Speaker 1: has an abysmal record on gender and racial equality. So 230 00:12:21,960 --> 00:12:25,200 Speaker 1: you know, having done this company, what what kinds of 231 00:12:25,240 --> 00:12:27,880 Speaker 1: things can be put in place that will change that 232 00:12:28,400 --> 00:12:32,320 Speaker 1: or improve it? Yeah, I would say you're you're not wrong, 233 00:12:32,559 --> 00:12:35,400 Speaker 1: and I think on the history of Silicon Valley on 234 00:12:35,480 --> 00:12:39,480 Speaker 1: the the silver lining, however, is that it is impossible 235 00:12:39,520 --> 00:12:42,320 Speaker 1: to have a conversation in Silicon Value where this topic 236 00:12:42,400 --> 00:12:46,160 Speaker 1: does not come up. Um, And I think, uh, we 237 00:12:46,320 --> 00:12:48,520 Speaker 1: and I say we, as somebody you know from from 238 00:12:48,559 --> 00:12:51,160 Speaker 1: that area, are very aware of this issue, and I 239 00:12:51,240 --> 00:12:55,160 Speaker 1: think that's probably one of the most important steps to take. 240 00:12:55,280 --> 00:12:58,480 Speaker 1: And so, um, I think we're trending in the right direction, 241 00:12:58,520 --> 00:13:00,120 Speaker 1: but I think we also have a long way to go. 242 00:13:00,320 --> 00:13:03,120 Speaker 1: So do you think real policy initiatives will follow that 243 00:13:03,200 --> 00:13:07,239 Speaker 1: greater awareness? Now, it depends what you mean by policy initiatives. 244 00:13:07,400 --> 00:13:10,440 Speaker 1: And if we're talking about I think a massive problem 245 00:13:10,679 --> 00:13:15,840 Speaker 1: like gender equality is going to take probably policy changes, 246 00:13:15,960 --> 00:13:19,560 Speaker 1: just awareness changes, cultural changes. There's a generational aspect to 247 00:13:19,600 --> 00:13:21,400 Speaker 1: it as well. There's a lot of things that have 248 00:13:21,520 --> 00:13:25,600 Speaker 1: to happen, and you know, regulation is one part of that. 249 00:13:26,160 --> 00:13:29,120 Speaker 1: So in two thousand fifteen, six years after you left, 250 00:13:29,200 --> 00:13:31,800 Speaker 1: you came back to read it as the CEO, and 251 00:13:31,840 --> 00:13:35,080 Speaker 1: you were soon confronted by a number of controversies, and 252 00:13:35,120 --> 00:13:36,560 Speaker 1: I want to talk about a couple of them. The 253 00:13:36,559 --> 00:13:39,840 Speaker 1: first was pizza Gate, and I don't know whether you 254 00:13:39,880 --> 00:13:42,679 Speaker 1: all are familiar with this crazy and false story, but 255 00:13:43,120 --> 00:13:48,439 Speaker 1: it started on Facebook and Twitter in October of UM. 256 00:13:48,480 --> 00:13:51,400 Speaker 1: The allegation basically was that Hillary and some people close 257 00:13:51,440 --> 00:13:55,160 Speaker 1: to her operated a child sex trafficking ring out of 258 00:13:55,200 --> 00:13:59,080 Speaker 1: a pizza parlor called Comet Pizza in d C. Which 259 00:13:59,080 --> 00:14:03,280 Speaker 1: would be laughable, um, except that the sort of fake 260 00:14:03,360 --> 00:14:07,240 Speaker 1: news had real consequences when a man named Edgar Welch 261 00:14:07,360 --> 00:14:11,200 Speaker 1: from North Carolina showed up at Comet with an a 262 00:14:11,360 --> 00:14:15,200 Speaker 1: R fifteen to rescue the kids, the sex slaves who 263 00:14:15,240 --> 00:14:17,840 Speaker 1: were trapped in the basement, only to find that there 264 00:14:17,880 --> 00:14:21,360 Speaker 1: were no kids and no basement. Um. And this was 265 00:14:21,400 --> 00:14:25,120 Speaker 1: on Reddit to both before and after the election. So 266 00:14:25,280 --> 00:14:28,160 Speaker 1: the bottom line is it could have been a real tragedy. 267 00:14:28,320 --> 00:14:30,880 Speaker 1: Do you ever feel like, as the founder of this platform, 268 00:14:30,920 --> 00:14:33,440 Speaker 1: it's the CEO of this platform, Oh my god, I 269 00:14:33,800 --> 00:14:37,120 Speaker 1: created a monster. No, not quite in that dimention. You know. 270 00:14:37,120 --> 00:14:40,720 Speaker 1: Pizza Gate was the community that that we saw develop 271 00:14:40,760 --> 00:14:42,920 Speaker 1: on Reddit, and we banned it after about two weeks, 272 00:14:43,600 --> 00:14:47,080 Speaker 1: long before the the incident with the the gentleman in 273 00:14:47,120 --> 00:14:49,840 Speaker 1: the pizza parlor. Um. And we do watch these things 274 00:14:49,960 --> 00:14:53,800 Speaker 1: very closely because while I don't think we want to 275 00:14:53,800 --> 00:14:56,880 Speaker 1: be in a position where we say conspiracy theories are 276 00:14:56,920 --> 00:14:59,800 Speaker 1: not allowed, right, because sometimes conspiracy theories end up be 277 00:15:00,000 --> 00:15:03,160 Speaker 1: being something um, But we do watch, especially in these 278 00:15:03,160 --> 00:15:06,560 Speaker 1: fringe communities, um, where is the where is the conversation going, 279 00:15:06,560 --> 00:15:08,680 Speaker 1: what direction is it heading? And with that one in particular, 280 00:15:08,720 --> 00:15:10,920 Speaker 1: we saw it going in a bad place, and it 281 00:15:10,960 --> 00:15:14,000 Speaker 1: was also easily disproven, and so that's why we got 282 00:15:14,040 --> 00:15:15,360 Speaker 1: rid of them as quickly as we did, and I 283 00:15:15,400 --> 00:15:16,520 Speaker 1: think we were one of the first to do so. 284 00:15:16,720 --> 00:15:19,320 Speaker 1: But could something like this happen again on Reddit? You know, 285 00:15:19,520 --> 00:15:24,000 Speaker 1: obviously the human awareness of this is not infallible. So 286 00:15:24,760 --> 00:15:27,280 Speaker 1: do you worry that this kind of situation, this kind 287 00:15:27,280 --> 00:15:31,160 Speaker 1: of inform misinformation can get out of control and lead 288 00:15:31,200 --> 00:15:34,480 Speaker 1: to something that may have far worse consequences. I do 289 00:15:34,640 --> 00:15:37,280 Speaker 1: think this sort of thing can happen in any place 290 00:15:37,320 --> 00:15:42,720 Speaker 1: where people congregate and misinformation is uh, you know, it's 291 00:15:42,760 --> 00:15:44,760 Speaker 1: it's a real threat and it's a it's a real 292 00:15:44,840 --> 00:15:47,880 Speaker 1: thing that we all face on Reddit and and anywhere 293 00:15:47,880 --> 00:15:51,480 Speaker 1: else online or offline for that matter. And you know, 294 00:15:51,520 --> 00:15:53,880 Speaker 1: I think one of the you know, we will do 295 00:15:53,920 --> 00:15:55,960 Speaker 1: our our our best to do our partner, but I 296 00:15:56,000 --> 00:15:59,120 Speaker 1: also think there is a collective burden we all have 297 00:15:59,480 --> 00:16:01,960 Speaker 1: as a as a people to be skeptical and to 298 00:16:02,120 --> 00:16:06,520 Speaker 1: push back on these stories, UM because you know, as 299 00:16:06,600 --> 00:16:09,240 Speaker 1: a as a private company or even an individual, we're 300 00:16:09,240 --> 00:16:11,840 Speaker 1: not always going to know all of the facts or 301 00:16:11,880 --> 00:16:13,480 Speaker 1: be able to see everything. So we will do our 302 00:16:13,520 --> 00:16:15,720 Speaker 1: best on our platform. And and in fact, we have 303 00:16:16,440 --> 00:16:18,920 Speaker 1: a number of teams dedicated to this, you know, in 304 00:16:18,920 --> 00:16:21,000 Speaker 1: our Trust and Safety Team and our Anti Evil Team, 305 00:16:21,000 --> 00:16:22,920 Speaker 1: which is our Trust and Safety team, is a team 306 00:16:22,960 --> 00:16:28,040 Speaker 1: of people that enforce our content policies UM and and 307 00:16:28,040 --> 00:16:30,520 Speaker 1: and in these communities, whether the conspiracy theory or generally 308 00:16:30,560 --> 00:16:33,280 Speaker 1: badly behaving, they generally violate all of our policies. And 309 00:16:33,320 --> 00:16:36,960 Speaker 1: so so our policies, for context are no harassing or bullying, 310 00:16:37,360 --> 00:16:42,640 Speaker 1: no involuntary sexualization of miners or or adults, uh, no spam, 311 00:16:42,760 --> 00:16:46,960 Speaker 1: no inciting or glorifying violence. And so we look at 312 00:16:46,960 --> 00:16:50,400 Speaker 1: these things very closely and and pretty pretty consistently. And 313 00:16:50,400 --> 00:16:52,280 Speaker 1: then we also have another team called the Anti Evil Team, 314 00:16:52,280 --> 00:16:55,000 Speaker 1: which is a team of engineers that supports the former 315 00:16:55,000 --> 00:16:56,760 Speaker 1: team at scale and tries to do this in an 316 00:16:56,760 --> 00:17:00,480 Speaker 1: automated fashion. Um, but the operative word being tried. I mean, 317 00:17:00,560 --> 00:17:04,600 Speaker 1: how can you really police all the things that are 318 00:17:04,800 --> 00:17:06,960 Speaker 1: part of Reddit when I how many posts are put 319 00:17:07,000 --> 00:17:08,840 Speaker 1: up there a day, Steve, I have no idea how 320 00:17:08,880 --> 00:17:13,160 Speaker 1: many millions, millions and millions, So can you really patrol 321 00:17:13,320 --> 00:17:18,520 Speaker 1: and keep this kind of misinformation even worse off the platform? Well, 322 00:17:18,600 --> 00:17:20,879 Speaker 1: so another thing that's that's really important to us, and 323 00:17:20,880 --> 00:17:23,760 Speaker 1: I think that separates Reddit from our peers is our 324 00:17:23,840 --> 00:17:27,240 Speaker 1: structure of community moderation. So all of our communities are 325 00:17:27,280 --> 00:17:30,880 Speaker 1: created by moderators, and those are those are not employees 326 00:17:30,920 --> 00:17:33,040 Speaker 1: have read it there, just users like anybody else, and 327 00:17:33,080 --> 00:17:35,800 Speaker 1: their job is to set the rules and standards for 328 00:17:35,840 --> 00:17:38,840 Speaker 1: their community and enforce those standards. And of course the 329 00:17:38,840 --> 00:17:42,040 Speaker 1: communities themselves are submitting all of the content and curating 330 00:17:42,040 --> 00:17:45,320 Speaker 1: the content. So every the structure of Reddit is basically 331 00:17:45,680 --> 00:17:50,440 Speaker 1: there are over a hundred thousand communities. Within those communities, uh, 332 00:17:50,560 --> 00:17:53,880 Speaker 1: community members, users, you know, me and you anybody can 333 00:17:53,880 --> 00:17:56,760 Speaker 1: submit content and then every piece of content can be 334 00:17:56,840 --> 00:17:58,840 Speaker 1: voted up or down. And so what you see is 335 00:17:59,200 --> 00:18:01,080 Speaker 1: what is the most popular at any point in time. 336 00:18:01,240 --> 00:18:03,760 Speaker 1: But the download is also important. Right in there, that's 337 00:18:03,760 --> 00:18:07,240 Speaker 1: where the community sets the tone, removes content that's inappropriate. 338 00:18:07,920 --> 00:18:10,280 Speaker 1: And one of one of our beliefs is that the 339 00:18:10,320 --> 00:18:15,320 Speaker 1: only thing that scales with users is users. And so 340 00:18:16,000 --> 00:18:17,479 Speaker 1: one of the things we spent a lot a lot 341 00:18:17,520 --> 00:18:20,040 Speaker 1: of time thinking about is how do we push this 342 00:18:20,160 --> 00:18:23,160 Speaker 1: burden of enforcement into our user base so that they 343 00:18:23,200 --> 00:18:27,199 Speaker 1: have control over this, because there's no way that a 344 00:18:27,200 --> 00:18:30,800 Speaker 1: company of our size, which is around four people, or 345 00:18:30,840 --> 00:18:35,240 Speaker 1: any size for that matter, can police the entire internet. Um. However, 346 00:18:35,280 --> 00:18:37,199 Speaker 1: I do think as a as a as a people, 347 00:18:37,720 --> 00:18:39,760 Speaker 1: whether we're talking about our community or if we were 348 00:18:39,760 --> 00:18:42,320 Speaker 1: to kind of take this these challenges up a notch 349 00:18:42,359 --> 00:18:44,879 Speaker 1: and just talk about our society in general as a people, 350 00:18:45,040 --> 00:18:47,159 Speaker 1: we can do that, and and and and as I 351 00:18:47,160 --> 00:18:49,960 Speaker 1: think all of our burden to to police ourselves in 352 00:18:50,000 --> 00:18:52,040 Speaker 1: that matter. In fact, you you wrote, you wrote a 353 00:18:52,119 --> 00:18:54,480 Speaker 1: letter to users, and you say it, I believe the 354 00:18:54,480 --> 00:18:57,320 Speaker 1: biggest risk we face is our own ability to discern 355 00:18:57,400 --> 00:19:00,720 Speaker 1: reality from nonsense. And this is a burden we all bear. 356 00:19:00,840 --> 00:19:03,919 Speaker 1: But isn't that a bit of a cop out basically 357 00:19:03,960 --> 00:19:10,479 Speaker 1: saying this is the user's responsibility primarily. I think if 358 00:19:10,480 --> 00:19:13,080 Speaker 1: you were to interpret that as this is only the 359 00:19:13,119 --> 00:19:15,560 Speaker 1: user's responsibility, that would be a cop out. Um. I 360 00:19:15,600 --> 00:19:18,080 Speaker 1: think we all have rules to play here. I read. 361 00:19:18,320 --> 00:19:20,280 Speaker 1: The thing that I think we are the most effective 362 00:19:20,280 --> 00:19:23,159 Speaker 1: at is building the tools and providing the framework for 363 00:19:23,200 --> 00:19:25,679 Speaker 1: doing these things at scale. And of course, something that 364 00:19:25,720 --> 00:19:28,159 Speaker 1: we were talking about earlier, which is very much a 365 00:19:28,200 --> 00:19:31,480 Speaker 1: technology problem, which is ensuring that the the volume of 366 00:19:31,480 --> 00:19:34,800 Speaker 1: any particular viewpoint on Reddit, whether it's positive or negative, 367 00:19:34,840 --> 00:19:37,240 Speaker 1: true or false, is in proportion to the number of 368 00:19:37,240 --> 00:19:39,879 Speaker 1: people who are saying it. Um. You know, to not 369 00:19:39,960 --> 00:19:43,320 Speaker 1: let our our systems be abused by bots or what 370 00:19:43,359 --> 00:19:45,800 Speaker 1: we call vote manipulation people, you know, trying to steal 371 00:19:45,800 --> 00:19:48,040 Speaker 1: the conversation in the wrong way. That's something that we 372 00:19:48,359 --> 00:19:51,240 Speaker 1: spend quite a bit of time thinking about, because I 373 00:19:51,280 --> 00:19:55,119 Speaker 1: think we should keep in mind the this this the 374 00:19:55,200 --> 00:19:57,840 Speaker 1: scale at what we're talking about, which is um. If 375 00:19:57,840 --> 00:19:59,920 Speaker 1: I were to paint with a broad brush, I would 376 00:20:00,000 --> 00:20:05,000 Speaker 1: say about Reddit mirrors humanity and about of our communities 377 00:20:05,000 --> 00:20:10,000 Speaker 1: and users are wonderful, creative, collaborative, you know, interesting people. 378 00:20:10,240 --> 00:20:12,000 Speaker 1: And I would say the same thing about this people 379 00:20:12,040 --> 00:20:15,960 Speaker 1: in general, and one percent are are nasty and they're 380 00:20:16,000 --> 00:20:19,240 Speaker 1: up to no good. And that means if we do 381 00:20:19,280 --> 00:20:22,679 Speaker 1: our job on the technology side and and and enforcing 382 00:20:22,720 --> 00:20:27,280 Speaker 1: that that can set appropriate standards and has visibility into 383 00:20:27,320 --> 00:20:30,000 Speaker 1: what's going on, we can we can drown out the negative, 384 00:20:30,560 --> 00:20:32,880 Speaker 1: you know, marginalize the negative, and do it that way 385 00:20:33,000 --> 00:20:35,960 Speaker 1: when it comes to dealing with that nasty one percent, 386 00:20:36,080 --> 00:20:38,639 Speaker 1: I'm trying to understand how much of that content is 387 00:20:38,720 --> 00:20:42,760 Speaker 1: dealt with by non Reddit employees, community managers who are 388 00:20:42,800 --> 00:20:45,879 Speaker 1: enforcing the standards, and how much of it really is 389 00:20:45,920 --> 00:20:48,320 Speaker 1: on you and how much But do you have to 390 00:20:48,359 --> 00:20:51,320 Speaker 1: actively go and take down So for example, the year 391 00:20:51,440 --> 00:20:54,679 Speaker 1: you came back, I read that you had to remove 392 00:20:54,760 --> 00:20:58,679 Speaker 1: two sub reddits. One was called fat People Hate, the 393 00:20:58,760 --> 00:21:01,760 Speaker 1: other was called coon Town, which were by the way, 394 00:21:01,800 --> 00:21:04,959 Speaker 1: as repugnant as they sound. And then the hate speech, 395 00:21:05,119 --> 00:21:09,879 Speaker 1: you know, sub communities went down, and overall hate speech 396 00:21:09,920 --> 00:21:14,000 Speaker 1: went down on the platform because participants in those communities 397 00:21:14,040 --> 00:21:17,560 Speaker 1: either left the site or they changed their language. But 398 00:21:18,440 --> 00:21:22,320 Speaker 1: it wasn't the moderators or the non employees who did it. 399 00:21:22,320 --> 00:21:24,480 Speaker 1: It was you guys who had to do it. Yes 400 00:21:24,520 --> 00:21:28,000 Speaker 1: and no, So those communities were repugnant and and and 401 00:21:28,280 --> 00:21:31,200 Speaker 1: we took them down. We also did behind the scenes 402 00:21:31,760 --> 00:21:34,399 Speaker 1: a lot of work on the individuals in those communities 403 00:21:34,960 --> 00:21:36,720 Speaker 1: um on. One of the things we've gotten a lot 404 00:21:36,720 --> 00:21:40,160 Speaker 1: better at over the years is how proactive we are. 405 00:21:40,640 --> 00:21:44,280 Speaker 1: So during that era, so this is two thousand fifteen, 406 00:21:44,840 --> 00:21:48,640 Speaker 1: we were reactive in terms of individuals reports. So something 407 00:21:48,680 --> 00:21:51,840 Speaker 1: bad is happening on Reddit, user reports it, then we action. 408 00:21:52,359 --> 00:21:55,200 Speaker 1: Now we are Gosh, I don't know the exact number 409 00:21:55,240 --> 00:21:56,120 Speaker 1: off the top of my head, I want to say 410 00:21:56,119 --> 00:22:00,520 Speaker 1: somewhere between se proactive, which means we are finding uh, 411 00:22:00,560 --> 00:22:03,760 Speaker 1: instances of bad behavior before it even gets reported. How 412 00:22:03,760 --> 00:22:05,320 Speaker 1: do you do that? Is it? Is it like playing 413 00:22:05,320 --> 00:22:07,240 Speaker 1: a game of whack a mole or or do you 414 00:22:07,240 --> 00:22:14,080 Speaker 1: have like software algorithms that can detect yes, yes to both? Yes, yeah, yeah, 415 00:22:14,080 --> 00:22:16,679 Speaker 1: I mean that's the thing. It's there. There is an 416 00:22:16,720 --> 00:22:18,520 Speaker 1: element of an arms race, there's an element of whack 417 00:22:18,520 --> 00:22:21,359 Speaker 1: a mole. Um, there's an element of of of being 418 00:22:21,359 --> 00:22:23,960 Speaker 1: clever with with how we find things. Again, as I 419 00:22:24,000 --> 00:22:27,359 Speaker 1: was saying before, we're talking about a relatively small population 420 00:22:27,400 --> 00:22:31,000 Speaker 1: of users, and even in those communities, um well, the 421 00:22:31,480 --> 00:22:38,520 Speaker 1: names are deliberately offensive and their motivation was in many 422 00:22:38,560 --> 00:22:42,160 Speaker 1: respects to draw attention to waste our time to troll 423 00:22:42,280 --> 00:22:46,160 Speaker 1: us um, and it was it was more organized than organic. 424 00:22:47,119 --> 00:22:50,280 Speaker 1: And that behavior stands out like we we can detect 425 00:22:50,320 --> 00:22:53,880 Speaker 1: that with computers. Um. And and so when we remove 426 00:22:53,920 --> 00:22:56,480 Speaker 1: those communities, yes, that the general tone of rehddick got better, 427 00:22:56,520 --> 00:22:57,960 Speaker 1: but I would also point out to the fact that 428 00:22:58,960 --> 00:23:01,719 Speaker 1: we got a lot better at UH enforcement on an 429 00:23:01,720 --> 00:23:05,320 Speaker 1: individual level, and then other communities around reddits, our users 430 00:23:05,359 --> 00:23:07,920 Speaker 1: were doing the enforcement right. If you show up into 431 00:23:08,560 --> 00:23:10,560 Speaker 1: you know, ask Creddit, Ask Gredda is one of our 432 00:23:10,640 --> 00:23:12,880 Speaker 1: largest communities, and and and if you're new to read it, 433 00:23:12,760 --> 00:23:14,240 Speaker 1: it's it's a place where I wuld invite you to 434 00:23:14,240 --> 00:23:18,960 Speaker 1: get started. Um. People post thousands of questions every day 435 00:23:19,240 --> 00:23:23,280 Speaker 1: that are completely you know, irreverence. UM. The one one 436 00:23:23,280 --> 00:23:26,480 Speaker 1: I saw the other day was teachers of creddits, Um, 437 00:23:26,600 --> 00:23:29,360 Speaker 1: what happened to that weird kid in your class? Um? 438 00:23:29,400 --> 00:23:31,840 Speaker 1: And it's just like at this incredible discussion and perspective 439 00:23:31,880 --> 00:23:33,760 Speaker 1: that you wouldn't see in the stories that you wouldn't 440 00:23:33,760 --> 00:23:36,800 Speaker 1: see anywhere else online. UM. If somebody shows up into 441 00:23:36,840 --> 00:23:44,560 Speaker 1: that community, yeah, exactly right. If you show up in 442 00:23:44,560 --> 00:23:47,960 Speaker 1: that community and and and start saying, you know, nasty stuff, 443 00:23:48,280 --> 00:23:50,119 Speaker 1: you're gonna get You're gonna get crushed long before we 444 00:23:50,160 --> 00:23:52,360 Speaker 1: see it right, You're gonna get downloaded by normal users, 445 00:23:52,400 --> 00:23:55,240 Speaker 1: You're gonna get banned by moths, and and that will 446 00:23:55,280 --> 00:23:57,359 Speaker 1: be the end of it. So we all have roles 447 00:23:57,359 --> 00:24:00,520 Speaker 1: to play on this. And I'm not claiming pfection by 448 00:24:00,520 --> 00:24:04,679 Speaker 1: any means, UM, but I do, ah, I am proud 449 00:24:04,680 --> 00:24:07,320 Speaker 1: of the progress that we are making. I think, especially 450 00:24:07,400 --> 00:24:10,520 Speaker 1: confronting what are it's a little bit of a new 451 00:24:10,520 --> 00:24:12,520 Speaker 1: frontier and a new set of challenges for I'm curious 452 00:24:12,520 --> 00:24:15,399 Speaker 1: how many humans you have actually with the sole job 453 00:24:15,720 --> 00:24:19,879 Speaker 1: of really trying to manage all of this. Dozens on 454 00:24:19,920 --> 00:24:22,640 Speaker 1: our staff. So our staff is four hundred, about half engineer, 455 00:24:22,920 --> 00:24:26,159 Speaker 1: half engineering. UM. One of our largest teams collectively of 456 00:24:26,200 --> 00:24:28,080 Speaker 1: the community, Trust and Safety and ANI Evil Team make 457 00:24:28,160 --> 00:24:31,400 Speaker 1: up a significant portion of our employees. To Evil Team 458 00:24:31,640 --> 00:24:35,000 Speaker 1: INTI Evil, Yes, that's the engineering team that UM you 459 00:24:35,040 --> 00:24:37,159 Speaker 1: know works behind the scenes, does the stuff at scale. 460 00:24:37,480 --> 00:24:40,880 Speaker 1: But then of course it is I do think both 461 00:24:40,880 --> 00:24:44,239 Speaker 1: on Reddit and in the real world, UM, we need 462 00:24:44,280 --> 00:24:47,880 Speaker 1: to enable ourselves to to self govern and to self heal, 463 00:24:48,520 --> 00:24:52,040 Speaker 1: because we don't want the government deciding what you can 464 00:24:52,080 --> 00:24:54,080 Speaker 1: and cannot say, And I'm sure you don't want a 465 00:24:54,119 --> 00:24:57,040 Speaker 1: private company, let alone me deciding what you can and 466 00:24:57,040 --> 00:24:59,439 Speaker 1: cannot say, or can and cannot think. I think that's very, 467 00:24:59,520 --> 00:25:05,440 Speaker 1: very dangerou It's time to take a quick break. We'll 468 00:25:05,480 --> 00:25:09,120 Speaker 1: be back with Reddit co founder and CEO Steve Huffman. 469 00:25:09,640 --> 00:25:18,159 Speaker 1: That's right after this and now back to our conversation 470 00:25:18,240 --> 00:25:22,000 Speaker 1: with Reddit CEO Steve Huffman, live from the Aspen Ideas Festival. 471 00:25:23,240 --> 00:25:25,399 Speaker 1: I want to ask you about this other challenge that 472 00:25:25,520 --> 00:25:28,800 Speaker 1: you've had related to the presidential election, that was Russian 473 00:25:28,840 --> 00:25:32,560 Speaker 1: interference and the proliferation of fake news. And by fake news, 474 00:25:32,600 --> 00:25:36,719 Speaker 1: I mean false stories, not stories the president doesn't like. 475 00:25:37,200 --> 00:25:40,520 Speaker 1: The Internet Research Agency, which was aligned with the Kremlin, 476 00:25:40,600 --> 00:25:44,040 Speaker 1: famously put fake news on Facebook, but they also put 477 00:25:44,080 --> 00:25:46,880 Speaker 1: it on Reddit. Do you ever think as a result 478 00:25:46,920 --> 00:25:50,240 Speaker 1: of that I contributed to the election of Donald Trump? 479 00:25:50,800 --> 00:25:54,560 Speaker 1: So uh, not as a result of that, I mean 480 00:25:54,760 --> 00:25:58,840 Speaker 1: the two reddits. It is where grassroots movements start, um, 481 00:25:58,880 --> 00:26:01,640 Speaker 1: and we've seen it with our last three presidential elections. 482 00:26:01,720 --> 00:26:05,639 Speaker 1: We were very very early on Obama. Um, it's those 483 00:26:05,680 --> 00:26:08,639 Speaker 1: as you know, I'm obviously extremely bicyclesn't right at all time, 484 00:26:08,640 --> 00:26:10,800 Speaker 1: but it was the first place where I learned of him. 485 00:26:10,840 --> 00:26:13,679 Speaker 1: We were very early on Bernie Sanders, right, who's who 486 00:26:13,760 --> 00:26:18,480 Speaker 1: had a very vocal grassroots campaign in the US, and 487 00:26:18,520 --> 00:26:20,600 Speaker 1: we're also early on Trump, who also had that kind 488 00:26:20,600 --> 00:26:25,280 Speaker 1: of populist support. As it relates to Russia, we published 489 00:26:25,560 --> 00:26:28,000 Speaker 1: a couple of months ago everything we found. We've published 490 00:26:28,000 --> 00:26:30,520 Speaker 1: all of the accounts that we found, everything they posted, 491 00:26:30,720 --> 00:26:33,760 Speaker 1: and that's still online for for our users, for journalists, 492 00:26:33,800 --> 00:26:36,760 Speaker 1: for for the government to to investigate and see what's there. 493 00:26:36,880 --> 00:26:39,639 Speaker 1: And we've been fully cooperative, uh, you know, with the 494 00:26:39,680 --> 00:26:43,560 Speaker 1: government behind the scenes this entire time. What we found, though, 495 00:26:43,680 --> 00:26:47,680 Speaker 1: is the structure of Reddit's largely worked. Of the nine 496 00:26:48,080 --> 00:26:51,639 Speaker 1: or so accounts that we believe we're compromised, um, a 497 00:26:51,720 --> 00:26:55,119 Speaker 1: handful fewer than ten actually found any meaningful traction on Reddit. 498 00:26:55,400 --> 00:26:58,960 Speaker 1: The vast majority were banned by moderators long before they 499 00:26:59,000 --> 00:27:02,880 Speaker 1: said anything, you know, anything relevant. So there was some there, 500 00:27:02,920 --> 00:27:06,080 Speaker 1: but there wasn't a lot of there there. The challenge, however, 501 00:27:06,119 --> 00:27:08,960 Speaker 1: that we face and and I and I need to 502 00:27:09,000 --> 00:27:12,800 Speaker 1: continually remind people of this that when we talk about 503 00:27:13,119 --> 00:27:16,360 Speaker 1: the president, sixty million people voted for that guy. Those 504 00:27:16,359 --> 00:27:20,879 Speaker 1: weren't Russian boats. Those were real Americans, And um, but 505 00:27:21,560 --> 00:27:24,679 Speaker 1: is what ideas were the Russian bots putting into the 506 00:27:24,720 --> 00:27:27,199 Speaker 1: minds of those Americans. I'm not suggesting that they were 507 00:27:27,200 --> 00:27:31,600 Speaker 1: all manipulated by fake news. I'm simply saying, were they 508 00:27:31,600 --> 00:27:34,439 Speaker 1: all making their decision on the basis of facts as 509 00:27:34,480 --> 00:27:37,280 Speaker 1: opposed to fake news? As you know, it's hard to 510 00:27:37,320 --> 00:27:40,120 Speaker 1: say because I think fake news is a new word 511 00:27:40,119 --> 00:27:43,600 Speaker 1: for propaganda and propagandas as old as politics. Um and 512 00:27:43,920 --> 00:27:46,040 Speaker 1: and we've seen this, and we see this on our 513 00:27:46,080 --> 00:27:49,320 Speaker 1: own you know, cable news channels, which are effectively the 514 00:27:49,320 --> 00:27:53,120 Speaker 1: propaganda department for for various political parties in the US. Um, 515 00:27:53,119 --> 00:27:56,920 Speaker 1: this is a real issue we face. Um uh, not 516 00:27:57,000 --> 00:27:59,320 Speaker 1: just on Reddit, not just on on Twitter and Facebook. 517 00:27:59,359 --> 00:28:02,359 Speaker 1: But I think as a people, but Russian interference was 518 00:28:02,440 --> 00:28:06,639 Speaker 1: relatively unprecedented in this last election, So certainly you can't 519 00:28:06,680 --> 00:28:10,320 Speaker 1: really compare sort of what you might term propaganda as 520 00:28:10,320 --> 00:28:13,400 Speaker 1: a spousing beliefs that would be supportive a particular candidate 521 00:28:13,480 --> 00:28:17,520 Speaker 1: or party with a foreign country actually interfering in our 522 00:28:17,560 --> 00:28:20,439 Speaker 1: democratic process. I think the Russia thing is a real 523 00:28:20,480 --> 00:28:24,800 Speaker 1: problem on on reddits because of the nature of Reddit, 524 00:28:25,200 --> 00:28:29,440 Speaker 1: those accounts, those efforts largely fell flats. Let's let's talk 525 00:28:29,480 --> 00:28:32,360 Speaker 1: about that. You said you have done and you're doing 526 00:28:32,359 --> 00:28:37,360 Speaker 1: your best to identify and remove Russian bots, Russian propaganda. 527 00:28:37,520 --> 00:28:40,280 Speaker 1: You said you've removed nearly a thousand Russian trolls from 528 00:28:40,320 --> 00:28:43,560 Speaker 1: your site. Just mechanically, can you walk us through how 529 00:28:43,640 --> 00:28:46,120 Speaker 1: you did that? How do you identify who's a Russian troll? 530 00:28:46,160 --> 00:28:49,400 Speaker 1: How do you get that troll off of Reddit? Sure 531 00:28:49,600 --> 00:28:53,400 Speaker 1: there are I don't want to get into the super 532 00:28:53,400 --> 00:28:58,120 Speaker 1: specifics because these techniques are still valuable to us, UM, 533 00:28:58,240 --> 00:29:01,000 Speaker 1: but when they help us as an ation, if you 534 00:29:01,040 --> 00:29:03,080 Speaker 1: did share those technics, well, let me shure, let me 535 00:29:03,080 --> 00:29:05,960 Speaker 1: share at at a high level. So, UM, there is 536 00:29:06,760 --> 00:29:09,840 Speaker 1: there's information we get from from our peers, right accounts 537 00:29:09,840 --> 00:29:12,200 Speaker 1: that they found i p s that they found we 538 00:29:12,440 --> 00:29:14,240 Speaker 1: as a as a best practice of Reddit. We've done 539 00:29:14,240 --> 00:29:16,040 Speaker 1: this long before the election and still do it now. 540 00:29:16,480 --> 00:29:19,880 Speaker 1: UM have a variety of information sources. For example, there 541 00:29:19,880 --> 00:29:23,480 Speaker 1: are gray hat hackers that basically that we can get 542 00:29:23,480 --> 00:29:26,000 Speaker 1: information from on. These are compromised VPNs, These are i 543 00:29:26,040 --> 00:29:28,400 Speaker 1: p s areas we should start looking. UM. From there, 544 00:29:28,440 --> 00:29:30,640 Speaker 1: you can find a couple of accounts UM, And by 545 00:29:30,680 --> 00:29:33,480 Speaker 1: gray hat you mean they they would be willing to 546 00:29:33,520 --> 00:29:36,000 Speaker 1: work for both sides in a in a way that 547 00:29:36,080 --> 00:29:39,000 Speaker 1: a bad guy can acquire. Um, you know, i'd say 548 00:29:39,000 --> 00:29:42,480 Speaker 1: there's a data breach, right, Equifax is hacked, right, and 549 00:29:42,680 --> 00:29:45,560 Speaker 1: they have all this information. That information is getting sold 550 00:29:45,560 --> 00:29:47,880 Speaker 1: to somebody. Sometimes it's getting sold to the good guys 551 00:29:48,080 --> 00:29:52,160 Speaker 1: who can say which of our which who has been compromised? Right? 552 00:29:52,160 --> 00:29:53,640 Speaker 1: And so that's very common. Right. And you can go 553 00:29:53,640 --> 00:29:56,040 Speaker 1: all these websites to see has is my information in there? 554 00:29:56,080 --> 00:29:58,280 Speaker 1: Is my email information in there? Then we can look 555 00:29:58,320 --> 00:30:01,120 Speaker 1: at behavioral things, right, and and this and this is 556 00:30:01,160 --> 00:30:03,400 Speaker 1: these are techniques that again we've been doing for a 557 00:30:03,400 --> 00:30:05,080 Speaker 1: long time. I read it because we care about the 558 00:30:05,120 --> 00:30:08,320 Speaker 1: integrity of our systems. You know, cookie hygiene, right? Are 559 00:30:08,320 --> 00:30:11,320 Speaker 1: people sharing this account? Are people in multile geography sharing 560 00:30:11,360 --> 00:30:13,120 Speaker 1: this account? Those things start to stand out. And so 561 00:30:13,200 --> 00:30:16,120 Speaker 1: from there we can build out a weapon. So you know, 562 00:30:16,160 --> 00:30:18,480 Speaker 1: we do it piece by piece by piece, and we 563 00:30:18,520 --> 00:30:20,600 Speaker 1: grow this, We grow this stuff out. So the stuff 564 00:30:20,600 --> 00:30:22,440 Speaker 1: we shared a couple of months ago was everything we 565 00:30:22,480 --> 00:30:25,520 Speaker 1: have found so far. Um, we are by no means, 566 00:30:26,200 --> 00:30:28,840 Speaker 1: you know, done that. This is as I said before, 567 00:30:29,120 --> 00:30:33,960 Speaker 1: Reddit's primary role in this is ensuring that our systems, 568 00:30:34,120 --> 00:30:37,360 Speaker 1: the integrity of our systems, and so there is, you know, 569 00:30:37,520 --> 00:30:39,800 Speaker 1: as as we said earlier, in an arm serice element 570 00:30:39,880 --> 00:30:42,719 Speaker 1: to this, and we're constantly learning and the challenges are 571 00:30:42,720 --> 00:30:46,200 Speaker 1: constantly evolving. But it is an issue we take incredibly seriously, 572 00:30:46,240 --> 00:30:48,320 Speaker 1: both as reddit and and as Americans. I know that 573 00:30:48,440 --> 00:30:52,760 Speaker 1: American redditors who then promote or repeat or share the 574 00:30:52,840 --> 00:30:57,720 Speaker 1: misinformation that is put out there by Russian bots, that's 575 00:30:57,760 --> 00:31:00,520 Speaker 1: a thorny r issue, isn't it? And work difficult And 576 00:31:00,520 --> 00:31:03,880 Speaker 1: that's where you believe user responsibility and our ability to 577 00:31:04,080 --> 00:31:08,440 Speaker 1: discern fact from fiction or fake news from real news, 578 00:31:08,840 --> 00:31:10,840 Speaker 1: is that's when that comes into play. I mean, if 579 00:31:10,840 --> 00:31:14,880 Speaker 1: we're to make it even thornier, the president is repeating 580 00:31:14,920 --> 00:31:17,520 Speaker 1: that nonsense as well, and so I mean we have 581 00:31:17,560 --> 00:31:19,920 Speaker 1: to ask ourselves some serious questions like what does it 582 00:31:19,960 --> 00:31:24,560 Speaker 1: mean when the president is is is propagating you know, 583 00:31:24,720 --> 00:31:27,440 Speaker 1: lies and fake news or whatever. Yes, that burden is 584 00:31:27,520 --> 00:31:30,480 Speaker 1: on us, and I think we have a some serious 585 00:31:30,560 --> 00:31:34,120 Speaker 1: challenges in front of us. I'll fight what does it mean? Well, 586 00:31:34,880 --> 00:31:38,240 Speaker 1: I think one of the the the silver linings of 587 00:31:38,240 --> 00:31:40,080 Speaker 1: all of this, and I think of this what I 588 00:31:40,080 --> 00:31:43,959 Speaker 1: would call it a challenging time is that our immune 589 00:31:44,000 --> 00:31:47,080 Speaker 1: system is being tested, right, and in the same way 590 00:31:47,120 --> 00:31:49,600 Speaker 1: that your body gets a virus and you're sick for 591 00:31:49,640 --> 00:31:52,040 Speaker 1: a couple of weeks and you eventually developed antibodies and 592 00:31:52,040 --> 00:31:54,120 Speaker 1: can fight against the next time. I think we as 593 00:31:54,120 --> 00:31:57,440 Speaker 1: a people are going through the same thing. People care 594 00:31:57,560 --> 00:32:01,440 Speaker 1: about quality journalism again, um, people are more more skeptical 595 00:32:01,560 --> 00:32:03,960 Speaker 1: than they have been in the past. I think our 596 00:32:04,120 --> 00:32:07,840 Speaker 1: values are being tested. Um, your values don't mean anything 597 00:32:08,680 --> 00:32:11,800 Speaker 1: until you use them. And I would say I think 598 00:32:11,840 --> 00:32:15,720 Speaker 1: we as as as Americans, UM, haven't had to haven't 599 00:32:15,720 --> 00:32:18,520 Speaker 1: had to flex some of these these these these muscles 600 00:32:18,520 --> 00:32:21,520 Speaker 1: and in a little while and and so well, the 601 00:32:21,840 --> 00:32:26,240 Speaker 1: medicine doesn't taste good. I think we will come out 602 00:32:26,240 --> 00:32:28,280 Speaker 1: of this. I hope you know. I'm an optimist. I 603 00:32:28,280 --> 00:32:29,800 Speaker 1: think we will come out of it healthier on the 604 00:32:29,800 --> 00:32:33,640 Speaker 1: other side, with a more skeptical population, with higher quality journalism, 605 00:32:33,640 --> 00:32:36,239 Speaker 1: with higher standards of discourse. But it doesn't feel good 606 00:32:36,280 --> 00:32:39,560 Speaker 1: right now, so to too awkwardly continue the metaphor how 607 00:32:39,960 --> 00:32:44,160 Speaker 1: strong is the body politic in eighteen as opposed to 608 00:32:45,200 --> 00:32:48,800 Speaker 1: how vulnerable are we to all of these factors and 609 00:32:48,880 --> 00:32:51,800 Speaker 1: forces this year? I mean, I would I would UM, 610 00:32:51,840 --> 00:32:53,600 Speaker 1: I often look to Reddit as a leading indicator of 611 00:32:53,640 --> 00:32:56,760 Speaker 1: what's going to happen. So, uh, in the two thousand 612 00:32:56,800 --> 00:32:59,200 Speaker 1: and sixteen election, if you spend any time on Reddit, 613 00:32:59,520 --> 00:33:03,640 Speaker 1: Trump when in the election was not exactly a surprise. UM. 614 00:33:03,760 --> 00:33:06,840 Speaker 1: The engagement of the Trump support us on Reddit in 615 00:33:06,880 --> 00:33:09,680 Speaker 1: terms of voting on Reddit was five times higher than 616 00:33:09,720 --> 00:33:12,640 Speaker 1: that of the Hillary supporters. So when I look at 617 00:33:12,680 --> 00:33:17,400 Speaker 1: what's going on now, um on Reddit the midterms, there 618 00:33:17,440 --> 00:33:20,440 Speaker 1: are a number of communities galvanizing around the mid terms. 619 00:33:20,520 --> 00:33:23,479 Speaker 1: And what kind of communities? What are they called? Like 620 00:33:23,760 --> 00:33:26,600 Speaker 1: I think once it's called blue mid terms. There's probably 621 00:33:26,600 --> 00:33:29,880 Speaker 1: half a dozen or so mid terms two thousand and eighteen. Um, 622 00:33:30,160 --> 00:33:33,240 Speaker 1: how are they weighed in terms of red versus blue? 623 00:33:34,120 --> 00:33:37,400 Speaker 1: Almost entirely blue. All of these new communities are in 624 00:33:37,480 --> 00:33:41,760 Speaker 1: response to, hey, we we cannot let our guard down, right, 625 00:33:42,040 --> 00:33:45,000 Speaker 1: is we have to show up and actually vote. Um. 626 00:33:45,200 --> 00:33:47,000 Speaker 1: I think you know, even when you look at the 627 00:33:47,040 --> 00:33:49,560 Speaker 1: voter turnout, it's it's always it's always dismal. And I 628 00:33:49,600 --> 00:33:52,560 Speaker 1: think that's the that's the largest opportunity we have going forward, 629 00:33:52,640 --> 00:33:55,440 Speaker 1: is is taking all these people who now care, who 630 00:33:55,520 --> 00:33:58,760 Speaker 1: now know what's at stake, and getting them to the polls. 631 00:33:59,560 --> 00:34:08,200 Speaker 1: I have to ask you about Charlotte's Ville. Yeah, vote everyone. Um, 632 00:34:08,239 --> 00:34:10,600 Speaker 1: As we mentioned, we both went to the University of Virginia. 633 00:34:10,640 --> 00:34:13,200 Speaker 1: You graduated just a couple of years after I did. 634 00:34:15,080 --> 00:34:19,640 Speaker 1: What's so funny? Um? Now, on a serious note, I 635 00:34:19,719 --> 00:34:22,600 Speaker 1: was in smack in the middle of that white supremacist 636 00:34:22,719 --> 00:34:26,600 Speaker 1: rally in Charlottesville last August. And I know that was 637 00:34:27,000 --> 00:34:31,080 Speaker 1: a real epiphany for you, wasn't it. Um? That was 638 00:34:31,160 --> 00:34:32,880 Speaker 1: very difficult, right. You know, I'm with a school in 639 00:34:32,960 --> 00:34:35,120 Speaker 1: Charles Cille. I grew up about an hour from there. 640 00:34:35,120 --> 00:34:38,240 Speaker 1: That is that is very much my home. And it's 641 00:34:38,920 --> 00:34:43,160 Speaker 1: was incredibly painful to watch, and and I still get 642 00:34:43,200 --> 00:34:46,279 Speaker 1: emotional when I think about it, and and and for 643 00:34:46,400 --> 00:34:49,839 Speaker 1: us that's we actually we were in the process at 644 00:34:49,880 --> 00:34:54,080 Speaker 1: the time of finding our voice in our content policy. 645 00:34:54,280 --> 00:34:56,640 Speaker 1: We spent a lot of time debating these and choosing 646 00:34:56,640 --> 00:34:59,520 Speaker 1: the right words, and so we were in the process, um, 647 00:34:59,560 --> 00:35:02,840 Speaker 1: prior to Charlotsville, of how do we talk about violence 648 00:35:03,280 --> 00:35:06,600 Speaker 1: and inciting violence and what are the rules there? And 649 00:35:06,640 --> 00:35:09,680 Speaker 1: then in Charles still happened, and I actually had a 650 00:35:09,760 --> 00:35:12,360 Speaker 1: very emotional reaction with my team where I was just 651 00:35:12,440 --> 00:35:15,960 Speaker 1: like just go ban all those people. Right. We had 652 00:35:16,000 --> 00:35:18,239 Speaker 1: banned the all right communities I think a month or 653 00:35:18,239 --> 00:35:20,120 Speaker 1: two before, but there are still kind of some fringe 654 00:35:20,120 --> 00:35:22,520 Speaker 1: you ones, and I was like, just just ban them all. 655 00:35:23,120 --> 00:35:27,279 Speaker 1: And fortunately, you know, I'm not the sole decider, you know, 656 00:35:27,280 --> 00:35:30,640 Speaker 1: in these conversations, and we have, UM, we have actually 657 00:35:30,719 --> 00:35:32,520 Speaker 1: quite a large team. It's it's me, it's our our 658 00:35:32,520 --> 00:35:34,560 Speaker 1: GC who's in the audience, our head of policy, our 659 00:35:34,600 --> 00:35:38,240 Speaker 1: head of trust and safety, UM debate these issues. UM. 660 00:35:38,280 --> 00:35:41,040 Speaker 1: But Charleston, I think was was very clarifying for us 661 00:35:41,040 --> 00:35:43,400 Speaker 1: and probably the country about like this sort of behavior 662 00:35:43,840 --> 00:35:46,680 Speaker 1: should not be tolerated. And in fact, as a result, 663 00:35:46,760 --> 00:35:51,360 Speaker 1: you updated your content policy banning content that glorifies insights 664 00:35:51,480 --> 00:35:54,279 Speaker 1: or calls for violence against an individual, groups of people 665 00:35:54,320 --> 00:35:58,239 Speaker 1: are animals, after noting that the previous rules that you 666 00:35:58,320 --> 00:36:01,680 Speaker 1: had you all had establed wish had been too vague. 667 00:36:01,719 --> 00:36:06,439 Speaker 1: And I'm curious, have you seen any positive result from 668 00:36:06,719 --> 00:36:10,759 Speaker 1: strengthening that policy? Sure? I mean when the policies always 669 00:36:10,800 --> 00:36:12,880 Speaker 1: really important because it allows us to make decisions a 670 00:36:12,880 --> 00:36:17,080 Speaker 1: lot faster, right, And and we're often kind of in 671 00:36:17,080 --> 00:36:20,839 Speaker 1: this position where we're trying to reconcile what our policies 672 00:36:20,920 --> 00:36:23,640 Speaker 1: say and what our gut tells us to do, and 673 00:36:23,680 --> 00:36:27,040 Speaker 1: can we find a way to articulate what our instinct 674 00:36:27,080 --> 00:36:28,719 Speaker 1: is telling us to do in a way that can 675 00:36:28,760 --> 00:36:33,640 Speaker 1: be enforced consistently and fairly, because that is is very important, 676 00:36:33,719 --> 00:36:35,880 Speaker 1: especially for a platform like us that is home to 677 00:36:36,640 --> 00:36:41,640 Speaker 1: you know, views um of incredible diversity and consistency and 678 00:36:41,719 --> 00:36:44,439 Speaker 1: enforceability is really important. But once we find those words, 679 00:36:44,520 --> 00:36:46,960 Speaker 1: even if it's a couple of sentences, um, then we 680 00:36:46,960 --> 00:36:48,880 Speaker 1: can go ham on these things, right, and we can 681 00:36:48,880 --> 00:36:50,640 Speaker 1: be like you're done, You're done, You're done, You're done. 682 00:36:50,680 --> 00:36:52,680 Speaker 1: And that actually feels very good. But I guess the 683 00:36:52,719 --> 00:36:56,080 Speaker 1: big question, as you mentioned earlier, Steve, is how do 684 00:36:56,160 --> 00:37:01,400 Speaker 1: you control harmful speech without becoming a censor and violating 685 00:37:01,400 --> 00:37:04,600 Speaker 1: the First Amendment? I mean that they're there in lies 686 00:37:04,680 --> 00:37:07,480 Speaker 1: the rub right. Well, that is the challenge with the 687 00:37:07,480 --> 00:37:10,799 Speaker 1: First Amendment. It is it is one of the founding 688 00:37:11,840 --> 00:37:15,120 Speaker 1: principles you know, over this country, and it's it's incredibly 689 00:37:15,160 --> 00:37:19,879 Speaker 1: important to our democracy, the marketplace of ideas and specifically 690 00:37:19,920 --> 00:37:25,000 Speaker 1: supporting political speech and and and debates, and so the 691 00:37:25,040 --> 00:37:26,600 Speaker 1: way we think about it, and if you look at 692 00:37:26,600 --> 00:37:31,400 Speaker 1: our content policy, it's largely structured around behaviors and not beliefs. 693 00:37:32,080 --> 00:37:35,239 Speaker 1: Beliefs are very different, difficult to police, and you are 694 00:37:35,360 --> 00:37:37,760 Speaker 1: free to have whatever belief you like, both on Reddit 695 00:37:37,800 --> 00:37:40,400 Speaker 1: and the United States generally, but it's your behavior as 696 00:37:40,440 --> 00:37:42,120 Speaker 1: we care about. And just like free speech in the 697 00:37:42,160 --> 00:37:45,760 Speaker 1: United States, which does have restrictions, right there's one about 698 00:37:46,200 --> 00:37:49,520 Speaker 1: inciting violence, UM, yelling fire in a crowded theater, slander 699 00:37:49,600 --> 00:37:52,440 Speaker 1: and slander and libel and false advertising. There are, in 700 00:37:52,480 --> 00:37:56,120 Speaker 1: fact many restrictions, but political speech is very very important 701 00:37:56,120 --> 00:37:58,839 Speaker 1: and very critical. And on Reddit we try to take 702 00:37:58,880 --> 00:38:01,120 Speaker 1: our cues from the government the same way. So our 703 00:38:01,120 --> 00:38:04,080 Speaker 1: content policies are things I shared before. Harassing and bullying, 704 00:38:04,200 --> 00:38:08,440 Speaker 1: inciting violence, those are behaviors, UM, those are more objective. 705 00:38:09,160 --> 00:38:12,480 Speaker 1: Your your your beliefs, your political beliefs in particular, we 706 00:38:12,600 --> 00:38:15,520 Speaker 1: do our best to to keep that off limits. That 707 00:38:15,600 --> 00:38:18,680 Speaker 1: doesn't mean we're not tested from time to time. And 708 00:38:18,680 --> 00:38:20,520 Speaker 1: and one of the actually the big lessons I've learned 709 00:38:20,560 --> 00:38:23,160 Speaker 1: over the last few years is when I came back 710 00:38:23,160 --> 00:38:25,759 Speaker 1: to read, I actually wanted to draw a fine line 711 00:38:26,000 --> 00:38:28,160 Speaker 1: this is good behavior, this is bad behavior. I have 712 00:38:28,200 --> 00:38:31,520 Speaker 1: since learned that that is basically impossible, right, wherever that 713 00:38:31,560 --> 00:38:33,799 Speaker 1: line is, somebody's going to walk right up against it 714 00:38:33,840 --> 00:38:38,000 Speaker 1: and stick their nose over it. Instead, we accept that 715 00:38:38,040 --> 00:38:40,080 Speaker 1: there's some gray area there, and that's where we spend 716 00:38:40,080 --> 00:38:42,800 Speaker 1: our time. And so we we define our content policies 717 00:38:42,880 --> 00:38:45,800 Speaker 1: to define that the clear area, and then we spend 718 00:38:45,800 --> 00:38:47,400 Speaker 1: our time as a team and the gray errors and 719 00:38:47,440 --> 00:38:49,920 Speaker 1: figure out is this, you know, is this on the 720 00:38:50,000 --> 00:38:51,560 Speaker 1: right or wrong side of our content policy? Do we 721 00:38:51,560 --> 00:38:53,960 Speaker 1: need to adjust our policies? This is a new behavior, 722 00:38:54,520 --> 00:38:56,680 Speaker 1: and and so it's a constant evolution. How do you 723 00:38:56,719 --> 00:39:01,160 Speaker 1: think Facebook is doing handling the issue of fake news 724 00:39:01,160 --> 00:39:03,960 Speaker 1: and bots and privacy and all the stuff that's emerged 725 00:39:03,960 --> 00:39:06,880 Speaker 1: since the election. I'll give you I'll give you two answers. 726 00:39:07,160 --> 00:39:10,120 Speaker 1: Um One, I have some empathy for what Facebook is 727 00:39:10,120 --> 00:39:14,880 Speaker 1: going through, uh, in that I know they didn't build 728 00:39:14,960 --> 00:39:18,080 Speaker 1: that platform for abuse, right. I brought that platform for 729 00:39:18,080 --> 00:39:20,680 Speaker 1: for friends to connect, and they built an incredible business 730 00:39:20,680 --> 00:39:26,239 Speaker 1: on top of it. Um. You know in Zuckerberg's congressional 731 00:39:26,520 --> 00:39:29,040 Speaker 1: hearings talking about how AI is going to be the 732 00:39:29,080 --> 00:39:32,040 Speaker 1: solution to all of their problems, I I don't believe that. 733 00:39:32,440 --> 00:39:34,920 Speaker 1: I think AI can help. Computers can help. But as 734 00:39:34,920 --> 00:39:37,920 Speaker 1: I said before, this is a people problem, and people 735 00:39:37,960 --> 00:39:41,040 Speaker 1: do we humans share some of the burden. And one 736 00:39:41,040 --> 00:39:42,200 Speaker 1: of the sayings we've had it read it for a 737 00:39:42,239 --> 00:39:44,279 Speaker 1: long time is is like humans do the hard part. 738 00:39:45,440 --> 00:39:47,879 Speaker 1: I think as it relates to privacy, we are very 739 00:39:47,880 --> 00:39:51,160 Speaker 1: different companies, right privacy to read it, that is in 740 00:39:51,200 --> 00:39:53,640 Speaker 1: our d n A. We don't know your name, We 741 00:39:53,680 --> 00:39:56,560 Speaker 1: don't know your gender, your age, barely know your location. 742 00:39:56,600 --> 00:39:59,359 Speaker 1: Because of your you know your your computer tells us 743 00:40:00,200 --> 00:40:01,839 Speaker 1: we don't want to know those things. We think that's 744 00:40:01,880 --> 00:40:04,640 Speaker 1: really important. Both we don't want the burden of of 745 00:40:04,760 --> 00:40:07,960 Speaker 1: personal information. We're not building a business off of selling 746 00:40:08,040 --> 00:40:12,200 Speaker 1: your personal information. And we believe that people when they 747 00:40:12,280 --> 00:40:16,879 Speaker 1: detached from their real world identity and their names, can 748 00:40:16,920 --> 00:40:19,920 Speaker 1: actually be more authentic, They can be more true to themselves. 749 00:40:19,960 --> 00:40:23,239 Speaker 1: There are many facets of your personality, uh and you 750 00:40:23,320 --> 00:40:25,879 Speaker 1: as a human that are not encapsulated in your name. 751 00:40:26,120 --> 00:40:28,480 Speaker 1: Right on this stage, I'm Steve Huffman on the CEO 752 00:40:28,480 --> 00:40:31,000 Speaker 1: of Reddit, but there are many communities I spend time 753 00:40:31,000 --> 00:40:33,600 Speaker 1: on Reddit where I am not the CEO of Reddit. Right, 754 00:40:34,000 --> 00:40:39,279 Speaker 1: I'm not telling you and that's important. Um and and 755 00:40:39,760 --> 00:40:41,759 Speaker 1: because you can find it, because you don't want to 756 00:40:41,760 --> 00:40:45,480 Speaker 1: share everything right, you know. Uh. For for example, community 757 00:40:45,960 --> 00:40:48,000 Speaker 1: I I learned about a couple of days ago. So 758 00:40:48,200 --> 00:40:50,440 Speaker 1: we do these these moderator road shows where we go 759 00:40:50,440 --> 00:40:52,759 Speaker 1: around the country and we meet with our moderators to 760 00:40:52,800 --> 00:40:54,759 Speaker 1: say thank you for what you do and to put 761 00:40:55,160 --> 00:40:57,239 Speaker 1: faces and names to to to the screen names that 762 00:40:57,280 --> 00:40:59,680 Speaker 1: we've known for years. Um. I met this couple in London. 763 00:40:59,719 --> 00:41:02,480 Speaker 1: They were at an American and a British person that 764 00:41:02,480 --> 00:41:05,360 Speaker 1: that it got married. They started a community called still Trying, 765 00:41:05,560 --> 00:41:08,520 Speaker 1: and this is for couples that are trying to have 766 00:41:08,560 --> 00:41:10,880 Speaker 1: a baby but struggling with it, and it's a place 767 00:41:10,920 --> 00:41:14,080 Speaker 1: for them to to vent, to support each other, to 768 00:41:14,680 --> 00:41:16,839 Speaker 1: you know, to go through this difficult time with other 769 00:41:16,880 --> 00:41:20,640 Speaker 1: people like them. And you wouldn't find that on other platforms, 770 00:41:20,719 --> 00:41:22,879 Speaker 1: especially associated with your name. You wouldn't would go talk 771 00:41:22,880 --> 00:41:25,200 Speaker 1: about some of these really personal issues. And I think 772 00:41:25,239 --> 00:41:28,799 Speaker 1: that's that for for us is what when we talk 773 00:41:28,840 --> 00:41:32,480 Speaker 1: about are we are? We are we proud of reddits. 774 00:41:32,920 --> 00:41:36,520 Speaker 1: We are certainly extremely proud to have to have brought 775 00:41:36,560 --> 00:41:38,920 Speaker 1: this out of people and to have created this enriching 776 00:41:39,000 --> 00:41:41,080 Speaker 1: experience in a place where people can be themselves and 777 00:41:41,120 --> 00:41:44,680 Speaker 1: be vulnerable and be authentic, and so that's really that's 778 00:41:44,719 --> 00:41:47,040 Speaker 1: what we deal on day to day. And so so 779 00:41:47,239 --> 00:41:48,920 Speaker 1: you think we're going to get to the other side 780 00:41:48,920 --> 00:41:53,480 Speaker 1: and closing, there's so much information, a tsunami of information 781 00:41:53,640 --> 00:41:57,040 Speaker 1: that sort of floods us every day that isn't verified 782 00:41:57,120 --> 00:42:00,600 Speaker 1: or edited, in many cases, no fact checkers, and it's 783 00:42:00,680 --> 00:42:06,239 Speaker 1: designed in in many ways to foment outrage rather than understanding. 784 00:42:06,480 --> 00:42:10,960 Speaker 1: So right now where we are today, not on the 785 00:42:10,960 --> 00:42:14,120 Speaker 1: other side, if you're hoping we'll get is this good 786 00:42:14,239 --> 00:42:21,799 Speaker 1: or bad for our democracy? I am hesitant to make 787 00:42:23,360 --> 00:42:27,640 Speaker 1: quick classifications of what is a really complex issue. I 788 00:42:27,680 --> 00:42:31,440 Speaker 1: think that the fat for hundreds of years in the 789 00:42:31,560 --> 00:42:37,319 Speaker 1: US that people can say anything and debate anything, that 790 00:42:37,520 --> 00:42:40,759 Speaker 1: is the cornerstone of our democracy. The fact that we 791 00:42:40,800 --> 00:42:45,279 Speaker 1: live in a time where it's less of a debate right, 792 00:42:45,360 --> 00:42:49,440 Speaker 1: and and more of a low brow, quick to judge 793 00:42:49,800 --> 00:42:54,080 Speaker 1: label everything, you know, identity, apolitis all these things that 794 00:42:54,239 --> 00:42:57,560 Speaker 1: is bad for our democracy. And so how do we 795 00:42:57,600 --> 00:43:01,960 Speaker 1: get back to the marketplace of ideas, getting the most 796 00:43:02,040 --> 00:43:07,359 Speaker 1: value out of free expression without losing it? And that's 797 00:43:07,400 --> 00:43:09,279 Speaker 1: one of the things I'm most scared about when we 798 00:43:09,280 --> 00:43:12,200 Speaker 1: talk about fake news. How do we eliminate fake news 799 00:43:12,280 --> 00:43:16,840 Speaker 1: without eliminating free speech and discussion and and real debates. 800 00:43:17,080 --> 00:43:20,120 Speaker 1: I think that is a real challenge. It's to end 801 00:43:20,120 --> 00:43:22,279 Speaker 1: on a positive note. I think we do all have 802 00:43:22,360 --> 00:43:25,600 Speaker 1: a role to play. I think as people, skepticism is 803 00:43:25,600 --> 00:43:31,120 Speaker 1: really important, as as as journalists being that arbiter of 804 00:43:31,160 --> 00:43:34,000 Speaker 1: truth is plays a really important role. As a as 805 00:43:34,000 --> 00:43:38,160 Speaker 1: a technologist, ensuring that the conversation is is is fair 806 00:43:39,480 --> 00:43:42,440 Speaker 1: and authentic is important as well. So I think if 807 00:43:42,440 --> 00:43:45,880 Speaker 1: we all play our roles and do our jobs, that 808 00:43:45,920 --> 00:43:50,240 Speaker 1: we will get back to a healthier place. Well, Steve Kaufman, 809 00:43:50,320 --> 00:43:54,560 Speaker 1: we really appreciate your honest and thoughtful discussion of these 810 00:43:54,800 --> 00:44:05,920 Speaker 1: really challenging issues. Thank you so much for joining us today. So, Brian, 811 00:44:05,960 --> 00:44:10,839 Speaker 1: I thought that Steve Huffman was forthcoming, But obviously there 812 00:44:10,880 --> 00:44:13,600 Speaker 1: are a lot of challenges that make it a very 813 00:44:13,640 --> 00:44:18,880 Speaker 1: imperfect system when it comes to monitoring hateful, disgusting, gross, 814 00:44:19,160 --> 00:44:22,440 Speaker 1: or just fake content on REDDA, as it is on 815 00:44:22,480 --> 00:44:25,279 Speaker 1: many other social media platforms. And that's why he's trying 816 00:44:25,320 --> 00:44:27,239 Speaker 1: to make the case that the burden rest as much 817 00:44:27,280 --> 00:44:30,080 Speaker 1: with the users as with the platform itself. I'm just 818 00:44:30,120 --> 00:44:32,360 Speaker 1: not sure how much the public is going to accept 819 00:44:32,400 --> 00:44:36,120 Speaker 1: that because there are dangerous real life consequences to the 820 00:44:36,160 --> 00:44:39,840 Speaker 1: sort of speech that gets disseminated on social platforms like REDDA. 821 00:44:40,120 --> 00:44:43,520 Speaker 1: I'm not sure people have the time or really want 822 00:44:43,560 --> 00:44:47,759 Speaker 1: to put the effort in and really evaluating the credibility 823 00:44:47,800 --> 00:44:50,399 Speaker 1: of some of the information that they're seeing. I mean, 824 00:44:50,440 --> 00:44:54,120 Speaker 1: I think they should be skeptical, obviously, but is that 825 00:44:54,200 --> 00:44:56,720 Speaker 1: really their job? I mean, that's an open ended question. 826 00:44:57,000 --> 00:44:59,480 Speaker 1: And there's also a chicken and egg open ended question, 827 00:44:59,560 --> 00:45:02,920 Speaker 1: which is is is the hate speech caused by the 828 00:45:03,000 --> 00:45:06,680 Speaker 1: platforms that give them a forum or was the hate 829 00:45:06,719 --> 00:45:09,240 Speaker 1: speech always out there and has read it just bringing 830 00:45:09,280 --> 00:45:10,920 Speaker 1: it to the fore or giving it more of a 831 00:45:10,960 --> 00:45:13,279 Speaker 1: spotlight than otherwise it would have had. Well, I don't 832 00:45:13,320 --> 00:45:15,799 Speaker 1: know if it's b I worry about sort of the 833 00:45:15,840 --> 00:45:19,879 Speaker 1: contagious effect of hate speech and how it incites other 834 00:45:20,000 --> 00:45:22,880 Speaker 1: people to be hateful. In other words, it kind of 835 00:45:22,960 --> 00:45:28,600 Speaker 1: unleashes hate that's been sublimated in our psyches. I don't know. 836 00:45:28,600 --> 00:45:30,799 Speaker 1: I'm getting too heavy deep in real right now. But 837 00:45:30,920 --> 00:45:34,560 Speaker 1: it was a very very interesting conversation, and I appreciate 838 00:45:34,600 --> 00:45:37,840 Speaker 1: the fact that he's willing to to feel tough questions 839 00:45:37,880 --> 00:45:40,319 Speaker 1: about his platform and about his business, and at least 840 00:45:40,360 --> 00:45:42,440 Speaker 1: right now, we're a lot more aware of these issues 841 00:45:42,480 --> 00:45:44,719 Speaker 1: than we were two years ago. There's much more of 842 00:45:44,719 --> 00:45:48,480 Speaker 1: a focus on preventing fake news and hate speech from spreading, 843 00:45:48,480 --> 00:45:51,239 Speaker 1: and so maybe that will have a positive impact. Has 844 00:45:51,280 --> 00:45:54,279 Speaker 1: that for an attempt at a happy ending. Well, we'd 845 00:45:54,280 --> 00:45:57,080 Speaker 1: like to thank the Asthment Ideas Festival again for having 846 00:45:57,160 --> 00:46:00,400 Speaker 1: us special thanks to Kitty Boone and Zach say Lewis 847 00:46:00,680 --> 00:46:03,520 Speaker 1: for all of their help, and thanks to Gianna Palmer, 848 00:46:03,600 --> 00:46:06,439 Speaker 1: Nora Richie, and Jared O'Connell our team at Stitcher who 849 00:46:06,440 --> 00:46:09,239 Speaker 1: helped make this show happen week after week. We'd also 850 00:46:09,320 --> 00:46:12,120 Speaker 1: like to thank Betha Mas and Alison Bresnik who helped 851 00:46:12,160 --> 00:46:15,480 Speaker 1: with the pod from their posts over at Katie Current Media. Also, 852 00:46:15,640 --> 00:46:18,560 Speaker 1: I wanted to say quick thank you to Eliza from 853 00:46:18,760 --> 00:46:22,240 Speaker 1: Costas Costas from the University of Virginia who's now working 854 00:46:22,560 --> 00:46:25,160 Speaker 1: at the Aspen Institute, who helped make sure that we 855 00:46:25,280 --> 00:46:27,440 Speaker 1: made it to the studio and of course that Brian 856 00:46:27,560 --> 00:46:30,799 Speaker 1: found the bathroom anyway, and she was a wonderful first 857 00:46:30,880 --> 00:46:34,200 Speaker 1: name by the way, right just like your little girl. Anyway, 858 00:46:34,239 --> 00:46:37,200 Speaker 1: Brian and I are the executive producers around here. Mark 859 00:46:37,200 --> 00:46:40,200 Speaker 1: Phillips wrote our theme music. Have you subscribed to our 860 00:46:40,239 --> 00:46:43,520 Speaker 1: show yet? Shame on you if you haven't. Yes, we're 861 00:46:43,520 --> 00:46:47,040 Speaker 1: really into shaming people here. Have you left a rating 862 00:46:47,040 --> 00:46:50,799 Speaker 1: an Apple podcast? Shame on you for that too. We'd 863 00:46:50,800 --> 00:46:52,359 Speaker 1: love it if you would do both of these things. 864 00:46:52,360 --> 00:46:54,879 Speaker 1: It really helps us out. In the meantime, we would 865 00:46:54,920 --> 00:46:57,480 Speaker 1: like to say thank you very much for listening, and 866 00:46:57,560 --> 00:46:59,799 Speaker 1: we'll talk to you next week. Thank you very much.