 Welcome to the evening sessions at NO, and we have a first speaker that is Willowbrew here. She is talking about weaponized social. Willowbrew is currently working with Vulpine Blue, where she helps coordinate distributed teams, and she's previously worked with humanitarian and digital response and has a vast experience in those fields. What you'll be talking about today is something very interesting. We've always talked about protocols for code. We've always talked about what are the ethics for code, but have we talked about interactions? No. Now she's going to talk about weaponized social, a way for social interactions to be protocolized. Hey, thanks. Oh, a big round of applause before she begins. I'm especially excited because Meredith is here, and she's the other, actually probably lifted even more on this project than I did, so I'm excited that you're here. So stick figures. I posted the link to the Prezi, which is what we'll be using here on my Twitter, so if you want to follow along later, you can. I like stick figures. So I'm interested in this question of what does equality look like? And as we do more and more things on the internet, we can look at more and more of our interactions. And we've always had access to some data, but there are questions of is equality in our access to resources? Is equality in the incidence of violence or lack thereof towards people? See, there's a little bruise on this one now. Is it in the way that we speak to each other? And it's in all of these things, but how do we know if we're getting there? So right now we know we have a lot of inequality in the world, and it's in things like where the gap in gender wages comes from. And I just read a fascinating book. If you or anyone you know is of the mindset of women do different sorts of work and it's not worth the same amount or whatever else, the history of marriage talks about how people used to get married as a business transaction, and you could love a lot of different people, but love inside the marriage was actually not a super good thing to do. And then in those business transactions, the women did certain kinds of work most of the time, but as we shifted into capitalism, the roles of the man going outside of the house were the ones that got more and more paid, and so the work of the women was less and less valued in this new system of economy. So it's really interesting, but if you want to know about the history of inequality, we should read more history. There's also inequality in where violence happens. So this might be on an individual level, or it might be at a more systemic level. I'm bringing these up because there are patterns in inequality, which is what we need to start looking at in order to address them, right? We need to know where to focus our energies in order to have the positive impacts on the world that we would like. So this is intimate partner violence. It's one person against another, but we also have systemic violence. This is the... I don't like this graph for a number of reasons. One is that they have named Latino as a race, Hispanic as a race, and that's your ability to speak Spanish, you're of a Spanish-speaking people, but that's fine. But we have this over-representation of non-whites in the prison system, right? And liberals built the prison system as it is now, and now it is systemically affecting some people more than others. We need to address that, right? We need to think about the patterns that we're a part of. So, how do we measure and increase and maintain equality? Lesig talks about four fulcrums for social change, markets, laws, social norms, and architecture. We like to talk a lot about markets. If it's good for us, they'll buy it, right? One of the difficulties of open-source software and Libre is dealing with markets and how we fit in, also how we pay ourselves as activists. Laws are not something I'm really going to get into here. I'm a terrible anarchist because I like law as the externalization of values of society. We expect people to adhere to certain ways of interacting. It's the enforcement part I'm not super keen on. But the thing that I want to get into here are the social norms and architecture. And while we can measure markets pretty easily because we know what people are buying and where money is going, we can track that more and more, and laws we're able to externalize and document, architecture and its effects on us and social norms are really difficult to measure. How would you do that? So while we have ways of measuring violence in some ways, not always, and we have ways of measuring income disparity as two examples, we don't have a way of really figuring out how social interactions are going. So there are these things called social scripts, which dates back in sociology to the study of human sexuality and the way that we know how to woo someone of our preferred gender if we're into wooing people is we've been surrounded by media and our parents and our friends, and we know that at least in the U.S. you might flirt a little bit and you try to get a phone number and you go out for coffee and then you hold hands and then you might kiss and there's this progression that we know about, right? But it's not the same in every culture in what is considered the most intimate changes. And so that's a script that we're following. So weaponized social, I'll get to that in just a second. Meredith put it incredibly well, so I'm just going to read this to you because that's terrible, is the existing harms of social scripts we ran while in smaller geographically constrained groups are being amplified due to network effects. Tiny unchecked errors scale become large harms as people find ways to exploit them in life just as in software. Does this make sense? Mesh, tiny nods? Okay. So we have a wiki, of course, because wikis, and there's a heart for Meredith because she's great. Okay, so we don't even have ways of understanding social interactions at the scale that we have now. We're running into all of these ethical issues around Facebook and other things because social scientists have never had data sets this big. Our sample sizes have always been so small that we struggle with statistical significance. And so our field has not progressed that much. So we don't have a lot of research or frameworks to work from in how to deal with things like d-dossing and doxing and other things in the social dynamics of that. Here's a proposed framework. It pulled it from a disaster and humanitarian response work, which is what I do most of the time. And there's a four-part cycle, which is preparedness. I live in San Francisco, so I have a go bag that waits by the door, so when there's the next earthquake, I have a decent pair of shoes and some extra clothes and some snacks and things like that. I am prepared, I have water. Response is what's going to happen after the earthquake or after the rubble or being pulled out myself. Recovery is after all of the aftershocks have passed, then we start rebuilding, right? And then mitigation is retrofitting buildings such that they don't collapse in the same way during an earthquake. It's changing the actual infrastructure so that the damage that is caused is different. So this is the framework that we've applied to this. I'll explain some more. So first, we have this thing where the extreme event in online space is not an earthquake. It's not the planet coming after us or us building in terrible places that we shouldn't live but are lovely like San Francisco. Hello, tiny spider. So on the wiki, we've broken down into these different components of what does the extreme event even look like. Then we get into preparedness and this is what a lot of InfoZack is based around, is how do you lock down your accounts, et cetera. So this was our first event in New York. It's Meredith and TQ on the laptop. Super fun. And so this is in how do you prepare against it? So how do you try to make safe space if that's a thing you want to do? How do you support people who are being activists? Do we even want to use codes of conduct? Like questions like that and if they're doing the work that we think that they should be doing. And also a personal reflection of how to critique me. Then we get into response. So the second event hosted was in Nairobi and we started thinking about how do people respond to online attacks or to large-scale social incidents online? So self-aware checklists, something bad is happening. I'm going to pause and think about if I'm actually making this worse or better, et cetera. Then we get into recovery. So what happens after social aspects have been weaponized? There's nothing here. We did a fair amount of research in a lot of different spaces and countries, et cetera. And if your online reputation gets sacked for something, there's no coming back from it. Like you change your name, right? And this is really concerning. If something goes poorly, we're not taking care of our people after the fact. We should fix that. And then we have mitigation. We also had a session in San Francisco that focused on mitigation. We hadn't planned that it was going to follow this cycle. It just kind of happened that way, which I find really fascinating. But we didn't take any pictures in San Francisco because we were so excited to build things. And I'm really proud of this work because it starts getting into how do we change the ways that we interact online such that these social harms are reduced but are not quashing anyone's ability or desire to speak. Does that make sense? Because it's not a choice between what you can say and who you can say it to. You should be able to say what you want to say. The freedom of expression does not have to be at odds with people's sense of safety and ability to participate. So we're talking about shifting social norms through architecture. Oh, I'm going to tell you about one of my favorite Slack channels that I'm a part of, which this is the graph of users added. So female-identified people are on the bottom, and so we've pushed to have gender equity and we've done a little bit better over time. So this is another version of that graph. But even though it's still slightly overrepresented for men, they do most of the talking. And so we notice this pattern. Like, what can we change about the expectations of communication in order to get these a little bit better? And we have... We're working on a couple of different options, but I want to tell you a story about how... No, I'm going to get there in a second. So here are things that we could be measuring in order to see if equality is happening in our communities. What does that mean? And are these the numbers that we want to be tracking? So we have the number of connections that someone can have, the threading model that you end up in, how different aspects of a thread are prioritized. Oh, no, it disappeared. Privacy control. So like, who can see what it is you're saying? Scope control. There are a lot of different aspects of this. Q, I think this was your part of the project that was amazing. And so we started outlining what are all of these different aspects of communication, and do we want to tweak these? Like, do we want to take a scientific approach to reaching equality where if we don't have representation happening the way we want it to, can we change our infrastructure in an explicit way and see if it actually impacts the way people are participating or not? Does that make sense? Like, let's science this shit. Yeah. So I'm going to tell you a story, though, about something that isn't measurable right now. My favorite channel in that Slack group that I told you about is called Awkward Silence. And someone created it in October of, like, last year and invited a bunch of people and no one said anything in it for, like, four months. And it was my favorite. And then one day I was like, oh, do you know what's going to be so good? And I posted the elephant emoji in it, and someone else immediately deleted it and changed it to the subject of the channel. And, like, that's how I like to participate in those spaces. It's not in typing messages, which is the thing that Slack tracks right now. You can track who says how much stuff, but I really like emojis and I really like moving text around and posting pictures and things like that. And so what we choose to measure is what we get is one of the systems theory things. So what are we choosing to measure in trying to reach more equality? Does that make sense? But making explicit choices in the things that you build so that people can play. So I was talking to Jonathan Stray about this talk and he brought up this excellent point that idealist design systems and pragmatists find paths. And I really like designing systems, but I'm also a person who posts the elephant emoji, right? And so even the systems that we design and that we're so pleased with and excited about to share, if they're not having the desired impact, we need to reassess, like find your path as other people also find their paths through your system. So this is me. I really didn't talk for very long. We should have a conversation. Do people have questions? I hope you have questions. All right. So this is me on Twitter. This is me on Mastodon. I hope it works. It's only going to work if we all use it. And major hearts again. And there are other people doing excellent work in this space. Sorry, I'll get there in just a second. So these are all wonderful. And again, this has already been posted to the internet. And if you want any help in designing and finding paths for this sort of thing, I am working with various groups on how to do that. Question, yeah. Does it work? Yeah, it works. Thanks first off for sharing your insights. I was wondering if what you and we are looking for, if there is a universal way to solve this, or if there should be multiple answers. I'm not even sure if answers is the right word. But to address this, is there one silver bullet? That's an excellent question. So this is one of the... We did this work in 2014 and 2015 mostly. And I had... And I think everyone did. But I had a really rough time because I was spending a lot of energy in online harassment groups. Just like how do we stop online harassment? And most of the women that I was working with there thought that it was a gender problem. Like it was a gender-based problem. It was violence against women. Not their social patterns. We run into the same sorts of attacks against religions, against races, et cetera, in other countries and in other cultures. I don't think there's a silver bullet. I think that there are going to be patterns that we find as we go through this space. I would love to do a large-scale analysis of what are the different parameters of interaction across different social platforms? How happy are their communities? And are they reaching representation? If representation is something that they care about. And basically offering a scientifically backed toolkit to people to make choices. Thank you. So I could imagine that what is universal is awareness, right? Awareness of challenges that are out there. But if you're talking about a framework, are we not already in advance to make some boundaries that other people in other cultures might misinterpret or be offended by? How could we address this sort of bootstrapping problem? Because people in other countries have been involved in the process of this project, I'm a little bit less concerned about that. And this is also not authoritarian at all. It's like, hey, we're taking this approach and you might find some use in it. But I have no desire to tell people how to interact or what to say or how to design their platforms. If you also share the goal of equality, maybe you should be taking deliberate steps towards reaching that. And here are some potential ways to get there. Thanks. So for the audience, I think I heard you as... Meredith's goal has always been exploratory. That list of knobs that you can twiddle in these online discussions, is it public anywhere? Yeah, it's on the wiki. Which wiki? The classic phrase. It's at weaponizesocial.aspirationtech.org. I'll post a link to it on both Mastodon and Twitter. Or you can email me. And we'll probably be migrating it to its own home soon, but we'll leave up redirects, etc. Thank you. Yeah. Hi. Hello. Can you... Lean into the mic. Can you comment whether moderation and evaluation of comments type functions have any impact on the quality and equality of the conversation? Yeah, so Nate Matias is actually the person who is the expert in that. He looks at community-based moderation. So not having an outside party be your moderator, but someone within your group. And his research points to it absolutely having an impact on it, but it has various impacts depending on how you're approaching it. He just polishes dissertation and it's under open access, so it's worth checking out. Thanks for that and always enjoy your presentation slides and the way you go about doing that. I was a little sad to see and hear that there's nothing really to do on the response side. And I guess there's so many different ways to go there, but maybe just because we're here from a technological or sort of architectural, you know, pie in the sky, are there things that you would sort of say are either traits or principles or even maybe specific examples such as like mastodon or other platforms that already do exist where you think a response is possible, or is that just not ever going to happen? I think that it must happen, so we'll figure it out. The only thing that has really come to mind so far is very pragmatic, which is a form of insurance. And so like what does it look like to have more resources at your disposal if something does happen to you? It's one of the reasons why I am one of the few people who actually saw some value in the right to forget thing and I know that it's abused in all sorts of ways and it's terribly executed and like I'm not saying that it's good as it is right now, but we don't have a mechanism for group forgiveness right now and it used to be that if you were, which happened rarely, ostracized from a small town, you could move to a different small town and now it's the internet and there is no other place to move to. And so our patterns of mass shaming are not sustainable and I am worried about how that's going to play out in the long run. Mike and then Merida, please. I was wondering, I mean, what are the representation of suppressed groups? Have you thought about engaging them because some of, let's say, even the open discussion in some cultures could be limiting? Yeah. Because it's not part of the culture or there's actually repression? Yeah, absolutely. But those are the same sorts of patterns of what we're reliving on the internet now. And so having, as I do my other work in helping affected populations, populations affected by a disaster or a humanitarian crisis, organize themselves and decide whether or not to interact with state actors or other agencies or international aid of some kind. This framework has provided useful, but there are definitely different assumptions about privacy and scope and of other things that are in that long list. And so having a baseline understanding of those tools and how it might impact a group or having them design their own tools I think would be key. I'm thinking about the risk of this project talking about these people rather than with them. But I have talked with them. How do you, for instance, address people who are women who are oppressed through sexual violence which is part of the culture to open up and to be part of the discussion? So, for instance, we use the same list of parameters when setting up a, even though we hadn't listed them out yet, we're reflecting back for a call center for victims of gender-based violence in Port-au-Prince. The same sorts of things apply to who should be able to look at this information, who is providing it, are they making connections to each other, all those parameters of interaction, and it's similar to helping design the SMS platform for women in Afghanistan to self-organize to tell stories. It's still the same sorts of things, but you have to work with the people who are going to use the tool, just like always. Okay. Thanks. Meredith. Going back to the topic of recovery for a second, one of the things that I used to fight with Casper Bowden about all the time was the right to be forgotten. He was in favor of it, I was against it. Until Andrea pointed out to me that it's basically digital exit. Talking about how online you can basically never hide from your reputation again, but that's one of the things that the right to be forgotten I think is intended to try to re-establish, although the teeth it actually has seem to be pretty limited. Yeah, I absolutely agree. To get to the earlier question about is there a silver bullet? There isn't. We need to be approaching this from multiple vectors at the same time, and so we should all be more forgiving, and we should all not escalate quite so quickly, but when there is a persistent problem, maybe we should try to get that person help and, like, what does justice look like and what does restorative justice look like, et cetera. But at the same time, until we have all of that sorted out, we also need to be thinking about how do we help people who have gone through hell, whether they deserved it or not. Like, they're still a human being and they need to be able to make a living after this sort of thing has happened to them. Right, and, you know, they still need to be recognized as a human being. Yeah. Exactly. Yeah, I mean, somebody else take it. That's good. I've been crunching a bit further on what my takeaways would be, and I was wondering, perhaps if Mark Zuckerberg were here and you were to give him one advice which he promised to follow up on, what would that be? I don't know that it would be a change in design in Facebook. To change the design in Facebook, I think it would be to give more autonomy to the users in what choices they make and the associated transparency to the algorithms necessary to make those choices. So having access to all of these toggles, but also being able to see the effects of changing those toggles on the algorithms, but being able to see it for yourself. Off the cuff, that's what I would ask. So more autonomy to individual users? Yeah, so there was the Facebook algorithm talk this morning that was fascinating. One of the questions was, if Facebook is running algorithmic analysis of images for nudity, why can't they do it for snakes, for someone who has a snake phobia? I do believe that we should be able to create our own experiences, but one of the things that's over here, I'm not sure where on the list it is, that did not do what I wanted to do. One of these over here is randomization slash serendipity slash enforced bubble popping. And this is something that is... So the word homophily is birds of a feather flock together. That's what it means. And people really like being with people that are like them. We're at a conference right now because it's so much easier to be around people that we don't have to explain everything to. And that's why I really enjoy the queer feminist meetups. And it gets just like, oh, okay, this is much better. But at the same time, that's not stimulating. It turns into a sort of... It breeds creativity in one way, while also inhibiting it in other ways. And serendipity online is really fucking hard. And so that's why it's in this list, is because people should at least be making a conscious choice that they are hiding from people, or that they are not wanting to interact with those unlike them. Because maybe their focus isn't equal representation and equal voice, et cetera. And then I don't share values with them and I'm not going to help them, but they're a human being and they should be able to do that. I think I would agree with that. Next step, would you go as far as saying they should be legally obliged? No. Why? One, because Facebook is a platform. They're not a government. So, well... Are they legally obliged to pop bubbles, or are they legally obliged to do what? No, pop bubbles or implements in the architecture measures that force people or let people see outside their bubble. I don't think that that falls under the rule of law in any way, shape, or form. No, it doesn't, I guess, but do you think it ought to be? Because if we're saying that this is a priority, the question is how should we implement it? We as the builders of platforms can do this. I don't want to just get into this, this is fascinating and we can keep doing more later, but... Yeah, I think you can have an offline chat with her. Do we have any other questions? Okay. Okay, let's do a hypothetical example. So maybe that gives you the opportunity or not to illustrate how you would do things and how you would do things differently. There was just a Google engineer who threw a curveball into the general consensus of how things work. And I'm just seeing the shitstorm unfolding. How would things under your proposed infrastructure go and how would they go different from how they're going right now? So one thing would be that as... Google would have needed to be transparent about their current gender ratios and hiring, et cetera, and also what sorts of pay scales people are on to show whether or not they... Even if you don't start off in a really equitable place, you can show that you're moving there. And so that is one thing that would have happened and then the point that this person is making would either be moot or not, because it would be obvious if Google was working towards equality or if they were not. And the other aspect of it would be that has the person's name come out? Okay, so right now he's probably getting shamed and pedestaled by all sorts of people, right? And so if Twitter had changed some of the way that information propagates on it, this would have been a much more calm thing and maybe the name would not have spread quite so much. And this then gets into like how do we spread really bad information that needs to get out really fast, right? And I don't know how to balance these things. But maybe he... And after all of this blows over, the shitstorm blows over, he would have a way to be like, hey, everyone, ends up I was super wrong. I finally read all the research or even a small amount of the research and how different people's brains works or don't work and ends up I was wrong and the crowd would go, oh, cool, we're glad you learned a valuable lesson and then it would be done and he would still be employable. Okay, you're implying two things here. That he's unemployable afterwards. And second, it's apparently a foregone conclusion to you that he is completely wrong and the counter-question to that would be the first and the last sentences that that guy wrote is basically, he's not sure if he's right, but he wants to basically pop a bubble and he wants to question a few articles of faith. Okay. And that part of the message got kind of de-emphasized and the message that's contrary to the common way things are seen is very overemphasized and everybody goes down on that and for example, things like the Norwegian model where everybody in the country has their salary information online, that is, that's completely not discussed. Okay. I don't understand what, it's fine. Do we have any other questions or comments? Okay. Thank you, Willow. Thanks. Thank you.