Greg Dunlap 0:04 So our guest today is Jasmine Sun. Jasmine is a writer exploring social and information ecosystems. She's currently on leave from Stanford to research tech policy at Schmitt Futures, and she also runs Reboot, a newsletter and events series on technology, humanity and power. So welcome to the show, Jasmine. Jasmine Sun 0:34 It's great to be here. Greg Dunlap 0:36 So I found some articles of yours online on Medium and they largely relate to how cities or other physical communities relate back to online communities. How did you kind of get interested in that and start writing about it? Jasmine Sun 0:54 Yeah, for sure. So, I mean, I think my story starts the way that a lot of people do who are interested in online communities and communities in general. So I grew up using online communities at an age when I couldn't go out and go and meet people. So I had the computer instead. So since I was like, I don't know, like 10 or so I had lots of anonymous WordPress blogs. And I would comment on other people's blogs and share posts. I use like Reddit, when it was like pretty early, I use this old web forum for pet bird owners called Tail Feathers. Like they're just all these like weird niche online spaces that I got really into, because of course, I was stuck at home as a kid. And I didn't have much better to do. In terms of the transition to how did I start thinking about these online spaces in relation to cities and how they organize themselves? Well, I've always loved traveling cities. So I've always been fascinated by the magic of a city, why it's so exciting to like, be in the middle of it and to explore it even without actively doing anything. So when I got to college, I took several urban studies courses, I was really interested in that. And I just started to notice all these parallels between the concepts that I was learning in class, like, whether that's something like social infrastructure, or eyes on the street, and the online communities that I was continuing to be a part of. And of course, I'm not like the first person by any means. It's this parallel. There's so many people who have talked about it. But it did become something that I kept thinking about over and over. Greg Dunlap 2:25 Yeah, I mean, I've always been really into cities too. And one of the things that I feel like I've always noticed is that, cities have personalities, just like people do. Would you say that online communities are the same way? I feel like the groups of people you bring together in something influence it in ways that are hard to define sometimes? Jasmine Sun 2:48 Yeah, absolutely. I don't know if you've read that Paul Graham essay called Cities and Ambition. But he talks about, like, how you can define and I swear, it's not like all the other Paul Graham essays, like it's not about like, I don't know, product market fit, or whatever else he writes about. But this essay, it talks about how you can tell the personality of a city by eavesdropping. And like, in some cities, when you eavesdrop, everyone's talking about money, money, money, career. In some cities, people are talking about their friends and their family. Like you can just sit in a coffee shop and different places around the city. And you get a sense of the personality from the conversations around you. And I thought about doing something, like, how do you do that? Similarly, in online communities, like just eavesdropping, like lurking, I guess would be the word. Right? And that's how you get the sense of the personality. Greg Dunlap 3:38 That's, that's so true. I mean, I think when I first met my wife, she lived in San Francisco, and I lived in Portland, and I would come down to visit her and I would be working out of coffee shops. And it would just, I just could not stand the way that everyone around me was always talking about their startups. It drove me absolutely crazy. And it's like, I just never, I just never thought about that. It's so true. Jasmine Sun 4:01 Yeah, yeah. I mean, I go to school in Silicon Valley, so it's very different there. Yeah, sort of enjoying the rest I'm getting during quarantine being there being at home. Greg Dunlap 4:12 Oh, yeah, I can imagine. So you know, one of the things that you've written about in your essays is how cities kind of act as platforms for people. What kind of similarities do you see there? Jasmine Sun 4:27 Yeah, so I think that this core similarity between cities and social media platforms that I always think about is what I'm just going to call the platform content distinction. And what that is, there's like two groups involved in both the actual architecture and the use of the platform. So you have one group that's designing and building the infrastructure, that they're usually a small group, they're probably powerful and they're probably more homogenous, less diverse on social media. That's going to be like the Facebook executives or their product. A team who are deciding on the individual features and policies in a city, that's probably the planners, the architects, the legislators. And then you have a totally different group of people who are the users generating content, they're interacting with each other. They're creating stuff. And they're using the space that has been offered to them by the first group. And these people are also culture creators, whether that's intentional or not intentional. Without great users, the platform is not a place that people want to be, and the types of users will decide who wants to be there. And like, these are the citizens of a city, right? They're probably way more diverse than the people designing the platform. And while users as citizens are constrained by the laws around them, by the places that they can go or not go, I find that users and citizens are always pushing back against those constraints. People are not just subjects to their technological overlords, people are always asking for new features or finding ways to sort of work around rules and constraints and laws. Greg Dunlap 6:06 You give the example of Reddit as a sort of a city as platform thing. How do you see that relate back to cities? Jasmine Sun 6:18 Yeah, so I think with Reddit, Reddit is really interesting to me, I feel like people don't talk about it enough. When they talk about social media, people are always talking about like Facebook, Twitter, Instagram, but actually Reddit is, I don't know what it is now. But like, it's very consistently in the top 5 or 10 sites that people around the world use. What makes Reddit interesting to me is because it's mostly strangers, you're not interacting with people you know, most of the time, and also that there are very distinct communities in subreddits, that I see as sort of like neighborhoods. Some are private, some are public places, you can enter and exit, but like you're not actually confined to a particular community, you're not confined to a neighborhood. But rather you can move between them, you can be part of multiple, you can visit here, and then leave and then visit another one. And so that's why I think Reddit is really interesting to explore, and where oftentimes this analogy makes the most sense. Greg Dunlap 7:12 One of the things that you wrote about in in that article was this idea of contested spaces. Like you talked about how skateboarders are people who push back on the norms of a city to use it for their own for their own purposes, maybe not in a way that the city defined it. How do you see that surfacing in online communities? Jasmine Sun 7:37 Right, so because you have the platform content distinction, the users are often not happy or dissatisfied with what they've been given, or maybe they are totally happy with it. But it's just more fun to figure out how to be subversive. To like, try to skate in places you're not supposed to, or do things like that. And so for example, with Reddit, what I'm always fascinated by how with a very, very simple UI, like literally Reddit is just a collection of web forms. There's not a lot in the ways of other features that shape your experience. Communities and moderators have found quite unique ways to create new features or norms. Like, there's a subreddit called Change My View, and you're supposed to try to convince other people, to persuade them to change their original opinion. And if you successfully do that, you get this thing called a delta, which is basically like a token or an award that shows that you have convinced somebody. But Reddit doesn't actually have a system of awarding deltas, like this is something that like one group of people just decided that like, this would be good. But what Reddit does have is you can put little tags or badges next year name where you can write any amount of text. So the moderators now say that if somebody awards you a delta, they keep a running record of how many you've gotten next to your name, it appears only in that subreddit. So now you have this sort of like social reputation token within a particular community. That was something that Reddit never planned to create, but people have sort of innovated or imagined in some way or another. Greg Dunlap 9:09 No, I hadn't really ever thought about that. Some of the Reddit communities that I'm a part of, those things, I think they're called flair, some of them have really, really complicated systems for like identifying yourself in the flair of their community or whatever. If I think about other technology platforms, like I think that the hashtag kind of came out that way, too, like people just kind of finding a use case for something that wasn't there. And so they just created a standard around it that kind of stuck. Jasmine Sun 9:48 Exactly. There are a lot of ways in which users find things that the platform executives or the designers never thought of in the first place. And then there's this weird where it's like, as a people who designs a platform, it makes the decision do I want to actually institutionalize this thing that my users are demanding? Like people are always asking Twitter for an edit button, but like I don't know. I'm always like, oh, there's so many ways and edit button could go wrong. Greg Dunlap 10:17 Oh, yeah. Yeah. So another thing that's sort of related to this in I mean, it's all kind of related the same way. But you talk about you talk about social infrastructures. So like, why what what do you what do you mean, when you when you say that in the context of cities and online communities? Jasmine Sun 10:56 Yeah. So social infrastructure is a concept from sociologist Eric Klinenberg. He wrote a book called palaces for the people that pitches this topic. He's also done a great podcast on 99% Invisible about it. So social infrastructure, what it is, is it describes shared spaces, physical spaces that bring all kinds of people together. And the idea is that these spaces spur civic life. So things like parks, libraries, churches, community centers, even a busy sidewalk or a bus, I think, to a lesser extent. And like the point of these places is they're not necessarily about your closest family and friends. Rather, they're about establishing weak ties, and creating an environment where people who would not ordinarily choose to go meet each other can learn to live and interact alongside each other. And this is important because, well, there's a lot of reasons that's important. Like Eric starts the book with this really powerful example about a heatwave in Chicago in 1995, and 700 people or something like that died in this heat wave. Those who are most affected tended to be older people, people of color, and who didn't have local networks, in particular people who didn't have friends and family or even like acquaintances in the area. And so they didn't have others to stay with, people to bring them food, and like just nobody to lend a hand. And so like, that's where I start to think of things, like mutual aid projects are often where our neighborhood based, done by community organizers, where you'll go around the neighborhood during the pandemic and get groceries for people who are immunocompromised or older folks who are at risk to COVID. And so social infrastructures, talk things about like, where are the spaces that you established these acquaintance space connections, these weak tie connections that are not people you talk to every day, but in a time of support and survival? That's where you can go to. like, I think about how America is a lot less religious and less churchgoing than they were a century ago. I'm not saying that's a bad thing. There are like lots of reasons that could be a good phenomenon. But like, churches used to provide like, the emotional, the social, and the financial support to struggling people, even if you don't really know them. And I was like, Okay, well, where are we going to replace that? We still need the support, even if people choose that religion is not something they want to be in their lives. So then it's like, okay, we are in a pandemic, nobody's going to the library. Nobody's going to church. Where are people going to find these resources? Where will you get mutual aid or where were you even just get exposure to different kinds of people and learn to become tolerant. I think about kids all the time, like if you're not going to school, how do you learn to share your toys and work with people who are not the same as you? And I think it's very tough for digital places to replicate the qualities that make social infrastructure effective. Greg Dunlap 14:00 Yeah, I mean, for for one thing, it seems like social spaces are much more ... your encounters with people seem like they would be much more random, right? Whereas online spaces seemed like their membership is much more curated. Does that make sense? Jasmine Sun 14:18 Yeah, absolutely. So like, there's like interspace networks, like things like Reddit, or this pet bird forum, where like, people are going in kind of a transactional way, like you go to get advice on like, what do I feed my pet or like, I need to buy like a tool, which one's the best one on the market? And I like these transactions. Actually, I think they're totally fine when you're talking about like your neighbors, like asking your neighbor for some flour or eggs or whatever, because that can actually convert to future support. Whereas online, I think it's very difficult for relationships in one arena, say, a particular hobby forum or like gaming, like lots of people make friends and games that they just like, never meet again, because there's no way for that to convert. That way they can in real life. Greg Dunlap 15:02 Yeah. And then of course the the selection of people who are available to you to meet there are to some extent self selecting, like there are people who are interested in platform gaming, then that makes it so that they're the people who can afford to be interested in platform gaming. Right. And so it's like, it seems like it's much more you're not getting as diverse a set of people as you would in your neighborhood or at the park. I mean, to some extent, neighborhoods are like that, too. But it seems like you are forced to encounter a much broader set of people out in the real world in general. Jasmine Sun 15:39 Yeah, yeah, I think that's where the whole contact theory hypothesis comes in. The psychology concept that you can reduce prejudice if in group and out group members interact provided a certain set of conditions. That's about like, equal standing, respect for norms, cooperation, things like that. But yeah, like in online spaces, there's just much less of it. And even if you're in a diverse group, I would say, you oftentimes don't know, or there are incentives to not reveal your identity, right? Like, you are a woman or a black person, or transperson. On the internet. I can imagine a lot of environments where you don't want to show people that. And, yes, I understand why, like I don't really want to be advertising in a lot of spaces. I'm a woman, certainly say not in a lot of gaming spaces. But as a result, people don't get the benefits of contact theory and learning to interact with diverse people. Greg Dunlap 16:35 Another thing that you talked about that was really interesting to me, as far as how tech has affected social infrastructure, is this idea that it formalizes what were previously informal relationships. So you talk about how you used to have to know a person, like if you wanted work done on your house, if you needed someone to come and fix something, you would either know an individual or you reach out to your community to get an individual and that increases your level of you know, of bonding in your community. Whereas with everything being sort of app based now, it's sort of rips those chords away. Jasmine Sun 17:15 Yeah, exactly. And I think that's one side effect of the gig economy, things like Instacart, or Uber, or whatever, I don't know, there's even like Rover for dog sitting. There used to be a lot of things that incentivize people to make friends, I guess. So like, small and tiny, like, "Oh, we can still make friends, just with other people." But like, now, when you want to send your kids somewhere, before you would try to know the other parents, for example. Now, yes, you can, but there's also so many alternatives. When you want food, you can just order delivery. When you need to borrow something. And I feel like this is one of those things like one of these side effects that people just don't think about, because it feels so small, until it's gone. And then you're like, oh, shoot. And then there's the inequality stuff that you mentioned, which is all of these services cost money, quite a lot of money, like food delivery food is so so much more than just going to get it. And that's how you get things like this heat wave where everybody's confined to their homes, or like COVID, you can't go outside, who's gonna be able to afford to get delivery groceries, who has to rely on social connections to do it as a favor, and who doesn't have those social connections and can't afford it, and will just never get the resources that they need. Greg Dunlap 18:34 And it seems like it would work the other way too, like if you're somebody who say, cleans houses for a living. And you know, one of the ways that you build that business is through word of mouth and connecting with people who refer you to their friends and family, who refer you to their friends and family and etc. But by abstracting that behind an app, it seems like you lose the opportunity to build those kinds of networks in the same way. Jasmine Sun 19:01 Yeah, I think that makes a lot of sense. I don't have familiarity, like, I've never been in that position. But I could imagine that it is hard to lose control over your business. And yeah, in those kinds of ways, like when you have these marketplace apps, whether that's like Yelp or Uber, basically centralizes all of these small businesses and transactions onto a single platform. And now you just lose control. Like, I don't know whether I'm going to be on the front page of Uber Eats today, or I don't know whether I'm going to be buried behind all this stuff. And yeah, I mean, I would imagine this would have a pretty disruptive effect on people who had already figured out this is what I'm gonna do with my business. And this is what I know, works. And all of a sudden, they decide to demote us in the algorithm. Who knows why. Greg Dunlap 19:48 And, you know, to be clear, I don't have any evidence that that was happening,. It was just something that occurred to me as we were kind of talking about this that that could affect the people who are looking for those services. And just as much that people are providing them as well. Jasmine Sun 20:03 Yeah, yeah, I was trying to think of examples because like, even even something like I remember when Facebook was doing their pivot to video, and they told everybody like, "Hey guys, like video is where it's at, if you are a media business." I mean, there's like College Humor or something like that, like, you got to invest all this in video, we're going to give you so many views, and they would like pump up the view counts on videos that people are posting, right. And it turned out that these views never converted to websites, all the ad money went to Facebook, it just like wasn't helping these content creators. And they had basically relied completely on the information that Facebook gave them, information that was inflated and exaggerated to benefit Facebook. And then a lot of business sort of went under for that reason. And I could imagine the same thing happening with restaurants that have to rely on third party marketplaces. Greg Dunlap 20:54 Yeah, yeah, for sure. One thing that is a topic that we've touched on on the podcast a lot in the past is moderation methodologies and how you can moderate your communities and content moderation and stuff like that. And you talk about how people will often think of that as a way to soothe or calm or even out discussions in the in their communities. But you say that you see it as it often having more of a combative effect. Can you talk about how that happens, or what you've seen as far as that goes? Jasmine Sun 21:36 I think that content moderation is something where it often just depends who you are, right? Like the analogy I think I make in one of the blog posts is to policing. I know that it's not the same, like physical violence is very different. But as with the police, you have some people who are like, "Yes, the police are here, they keep us safe. I want more and more and more police in my community." And that group of people probably has more social power, like in our neighborhoods, they're probably white, and they're not the ones being targeted, right? So the same thing happens online where content moderation is especially tough when there are perceptions or actualities, that there is either bias or just a lack of transparency. This happens oftentimes, like Facebook, and all these other companies, they're moving towards trying to use algorithms more and more in content moderation, obviously, because of scale reasons, and also it sucks to make humans look at all this horrible content. But stuff gets taken down accidentally, like the AI is not that good yet. And some communities end up getting over-policed, like sex workers, activists often complain of getting over policed. So do people on the far right, they also often say our content gets policed way too much. I'm not gonna like adjudicate like which of these people have the correct claims, but the point is that people just don't know what standards are being applied to them. There's no explainability. With algorithms, it's very, very hard to get personalized responses. And like, even with the police, like body cameras, or reports that you have to write down where you can go back and review everything that has been done. Like, there's absolutely no way on a privately owned social media site to interrogate what kinds of social effects or biases are being replicated in the way that content is taken down. Researchers are trying, they have proxies, but it's just really hard. Greg Dunlap 23:32 Yeah, and I guess it's even harder when you're in a platform like Facebook where, you know, nobody really agreed up front to a, I don't know, worldview or you know, level of moderation or topicality that's acceptable. Facebook, and at least for a long time, Reddit, it seems to be changing a bit now, but, you know, seem to be much more billing themselves as like an open forum, rather and that's an issue because when you try and then later create norms around it with people who have already come in with a certain expectation, obviously, that causes a lot of turmoil. Jasmine Sun 24:15 Yeah, expectation mismatch is huge. And I think a lot of like, you're never gonna make everyone happy, but I do think a lot of it is things like communications and expectations management. And I feel like a lot of times users of platforms feel kind of tossed around by feature changes, algorithm changes, content, moderation policies, like all this stuff changes. You don't get any advance notice and all of a sudden, your social life, or even your livelihood, just gets sort of passed out the window. And so there's this feeling of distrust that these outsiders who have no understanding of who you are or what you want, just come in, impose their policies. And then you just feel targeted. Like I understand why this is seen as external forces. And I think maybe, and maybe you've talked about this in a lot of your other conversations, my guess is that more smaller scale, local moderation is a lot better. Like you see this in certain Facebook groups or like subreddits, when you come in, there are particular moderators who are part of the community that they are also moderating. So there's sense that they're one of us, they know us. And these people set rules that are particular to their community. And I feel like, yes, people can still get angry at the mods, but there tends to be less of that distrust and push back then with site level admin. Greg Dunlap 25:42 No, and I think that you're right, that a lot of it just comes from expectation management, right? Because when somebody comes into that community, they're presented with a set of rules that they can agree to or not before they start participating, as opposed to people who come in expecting something that's open, and then feeling like the rug gets pulled out from under them or whatever. Jasmine Sun 26:02 Yeah, yeah, exactly. Like I, I would never say that if you are a platform administrator, you should have no content policy, like everyone can just make whatever rules. Like if your a subreddit, wants to break out, break the law, like go for it, but I think it is about like, it would be better had a lot of these large platforms set early on a set of transparent rules and policies for like ... the Hunter Biden story on Twitter The people's problem, I don't think, was just that it was taken down, it was we have no idea why or like what policyt this was taken down under. Had these platforms been a little bit more proactive in setting out this is what we will allow, this is what we won't allow here, all our policies, we're gonna keep updating you guys as they change. I just think that that kind of consistency and transparency would just have a huge effect. Like in a city, for example, you could go look up the laws about what a hate crime is, you can go look up a database of all the hate crimes that have been reported or hate speech that has been reported. You can't do that on a platform. Greg Dunlap 27:04 Well, yeah, and certainly not on the big ones like Facebook or Twitter, whose content guidelines seem to be not just changing, but changing depending on who they get applied to? Jasmine Sun 27:17 Yes, yes, of course. Greg Dunlap 27:19 So who do you see doing kind of online social experiences well these days? You've talked about Reddit, but is there anybody else? Jasmine Sun 27:29 Well, first to clarify, I wouldn't say that like Reddit as a whole. Greg Dunlap 27:33 Okay, sure, right. Jasmine Sun 27:38 I attribute a lot of the things that are going well, in particular communities, to those moderators who are a part of their community and who figure something that works out for them. I think it's hard to say because I would say that a lot of people doing it well, are the the ones in who are managing smaller scale communities. And they're not the kinds of things that you see on the news, they're thousands of small scale communities across the internet, on all of the platforms that seem to have figured things out. I think it is just like what you said about when you have scale, when you say global connectedness, and you are attempting to unite millions and millions of people, billions of people in Facebook's case, under one platform instead of policies, you're always going to have tension. I will say that I think that it's good that Twitter is being more proactive. I just wish they were more consistent and transparent about their policies. But I like it better than the Facebook case. I think. Just not no policies that are very, I don't know, it's tough. Do you have any? Greg Dunlap 28:50 No, I mean, I was just gonna say that what you bring up that has been kind of a topic on several of my podcasts, which is that when you when you have a smaller or more niche community, your goal is community. Right? But when you have something at scale, your goal is profit. And I think that that's something that's a real dividing line between the people who are doing it well and the people who aren't because doing community when profit is your goal is very hard, I think. Jasmine Sun 29:25 Yeah, absolutely. Like that's, I think, where the whole platform is like a city idea really breaks down, which I think I mentioned briefly, but its a whole other thing that needs to be explored. Because in a city yes, there is a trust and a power gap between citizens and the government. But at least in much of the US, it's somewhat democratic. And there are ways to file public records, request complaints, there are elections where you can vote people out, there referenda where you can vote law in, and all these built in institutional incentives where there's a direct pathway for citizens and people to influence their government. Whereas with a for profit company no matter how many oversight boards they create, no matter how many committees are like, where you're gonna listen and learn to do a listening tour or whatever Zuckerberg did, it's not actually democratic, right? We all know this. They're accountable. They're accountable to shareholders. And, yes, they do user research, and they listen to users, but they only do that insofar as it's how do we keep users happy enough that we keep making money? And in particular, how do we keep the users who make money on this platform? Because like, not all users are valued the same? Greg Dunlap 30:43 No, I think all of that is absolutely 100%. True. So what other lessons do you think that online communities could take from civic infrastructure and urban planning to improve things for their members? Jasmine Sun 31:02 Yeah, so one thing that I mentioned a little bit is just you have to recognize that you are the decision maker, you are an authority. And then you've got to make those policies and ethical judgment calls. I'm not saying don't make them, but make them, say what they are, and stand by them. And if somebody changes your mind, also stand by that and be very communicative and transparent, manage those expectations around that, and I think that will just do a lot. I think about how, let's say you want to do a major new development in a neighborhood, and that could potentially be perceived as gentrification. And if you are a planner, or the city council representative, or the government interfacing with this community, there are going to be concerns. And before you actually just start building stuff, or approving business permits to build, you go and talk with members of the community, you ask them what they need, you actually respond to them, you change plans, you approve various benefits and stipends in reaction. And so I think, in the same way, when a company is about to make a major change or feature that might affect their users in different ways, these can be communicated better, they can be communicated before they're made, you can get user input along the way in public forums. And not only with particular focus groups that nobody knows about, but doing like things like impact assessments, especially on edge case users. Whether that means people who are activists, whether that means sex workers often get booted off platforms, and they lose their livelihoods that way, like all these users who have a lot more at risk. There needs to be impact assessments before implementing feature changes or new content moderation policies to understand how will they be affected. Again, I don't know whether people will do this, because there's not a huge incentive to keep some of the people on your platform if all you hear about is money. But I would like to see some of those community engagement practices implemented in social media platforms. Greg Dunlap 33:13 Cool. Well, I love this topic. And I wish we could go on about it for much longer than we have here today. If people want to learn more about your work or what you're working on these days, how can how can they find you? Or what are what else you working on? Jasmine Sun 33:33 Yeah, for sure. And thank you again, for inviting me on. I love talking and thinking about this stuff. If people want to find me, so I use Twitter pretty regularly. I'm at jasminewsun. I also run a newsletter and event series called Reboot on tech, humanity, and power. We do author events twice a month, we review the latest books in the genre, we publish original essays. If you want to subscribe, you can find that at reboothq.substack.com. So those are the main channels by which I push out writing and things like that. But yeah, thank you for giving me this opportunity to come and share. Greg Dunlap 34:11 No and thanks for coming on. I really appreciate you taking the time. Transcribed by https://otter.ai