I want to make a better social media site.

"Remember to imagine and craft the worlds you cannot live without, just as you dismantle the worlds you cannot live within." – Ruha Benjamin

For the past four years I've wanted to make a social media site alternative to Twitter (now X). I didn't have any hope in it becoming successful — and to be honest, I still don't. But at the very least, I want to share this post as a first draft of some of the ideas I've had, especially since folks have been asking quite a bit about it.

Quilt is the working name of the social media site. It's intended to be text-first, in the way that Twitter was, as opposed to Instagram and TikTok. I wanted to choose a name that evoked something cozy and connective, and this name started as a reference to a story quilt, or narrative quilting. With this name, I also wanted to acknowledge how social media sites allow people to collaboratively build stories, discourses, and narratives with each other, through literal and digital threads.

Quilt is intended to be consentful, safe, and joyful.

• Consentful social media. This is directly inspired by research from Jane Im, who proposed a design for social media based on the framework of affirmative consent. A lot of tech products aren't very consentful — they sell our data without informing us, and they put us in situations where we don't really have choice and agency over our interactions. But what if we imagined a social media site that was consent-first, that gave us more agency over how we engaged with our feeds, how we protect and private our content, and how we're interacted with by others on the site? At the core of Quilt, I want to explore this framework and design with consent at its foundations.
• Safe for marginalized communities. So many sites today are absolutely toxic and dangerous for marginalized folks across different identity axes. Whether it's dealing with anti-Blackness, misogyny, anti-queerness, transphobia, ableism, and so many other forms of hate and harassment, it feels like there aren't spaces that take marginalized folks' safety seriously. I'm tired of the idea that all social media platforms need to be for all people. This has led to content moderation and community guidelines that in practice end up supporting the dominant, status quo opinions, which are typically extremely hateful and bigoted. I don't think Quilt should be "neutral" (which is an illusory and flawed idea), but instead it should make a stand to be safer, particularly for marginalized community members. If folks don't like that, they can choose to use a different platform, like Twitter or Bluesky or Threads or any of the other options out there that defend hate speech as a form of free speech. I plan to adopt Afsaneh Rigot's approach of design from the margins, where we would center marginalized folks at all points in the creation and ongoing development of Quilt.
• Joyful or playful social media. To balance out the idea of safety, Quilt also necessarily emphasizes joyfulness or playfulness. I think there are a lot of takes on "safety" that end up with sites that are extremely sanitized and soulless. I don't want Quilt to be another professional-feeling, SFW-only space, especially when the term "safety" has been used to unfairly target queer folks and sex workers. While consentfulness means the ability for people to hide content that they don't want to see, consent doesn't mean we're taking an abstinence-only approach (which itself doesn't work). One of the reasons that researcher Andre Brock argues that Black Twitter really flourished was because it could be "ratchet," and Quilt should be a space that allows for that as well. Social media should be fun, and I really want us to go back to that.

Some more specifics:

• Better feeds, both chronological and algorithmic. In the last few years, there's been a huge trend of pushing back against the algorithmic feed in favor of a pure, reverse chronological feed (aka, a feed where content is sorted from newest to oldest). I've written in the past on how chronological feeds serve a specific purpose (such as if there are important live events occuring, you want to be able to see what people are saying in some ordered fashion), but they have some tremendous drawbacks as well (such as forcing users to have to constantly check the feed to stay updated, penalizing those who don't post as much, and making it hard to discover important or interesting content). It's not necessarily the case that algorithmic feeds are bad, and a group of researchers and technologists actually put together a report that shows how we can design better algorithmic feeds that better serve people using social media. Because Quilt is consentful, we still want people to have the option to choose whether they want the reverse chronological or algorithmic feed, because both have their pros and cons.
• Privacy. One of the things lacking from Bluesky (which I'll comment on later) is the ability to private your account. I think privacy is so extremely key to a social media platform, in terms of consent and safety, but also because sometimes people just want to be private on the Internet. Right now we're seeing an embrace of technologies that make things more open, but in a way that is dangerous for marginalized folks who want to be able to hide and protect themselves, especially from malicious actors, harassment, government targeting, data scrapers, etc., and Quilt needs to be a platform that allows people to feel safe being themselves and communicating with their communities. From private accounts to end-to-end encrypted DMs to the ability to make posts visible only to a select circle, I want Quilt to embrace privacy practices that give more control to the users and more protection especially to deal with hate and harassment.
• Data privacy and protections against Generative AI. Connected to privacy is the fact that I want Quilt to offer users data privacy. I don't want it to be a platform where users' data is sold off. Even if for many social media sites that's a huge way for them to make profit, I adamantly want Quilt to be funded by other means, and keep the data of the users safe and private. Along those lines, I want to provide as many protections against Generative AI as possible, such as building in the ability to optionally apply Nightshade and Glaze to images that are posted. I also want to design Quilt to be as helpful and friendly to artists as possible, such as finding ways to watermark screenshotting, or ways to ensure credit for artwork is attributed to the artists even as its being reposted.
• Transparent and contextual content moderation. In my opinion, one of the biggest issues with social media today is the lack of transparent and contextual content moderation. If we truly embrace designing from the margins, we need to first acknowledge that a lot of "automated" content moderation simply doesn't work and often silences marginalized folks, while invisibilizing the labor of actual human content moderators that are doing the work (linked: the work of researcher Sarah Roberts who extensively studies the outsourcing of commercial content moderation). In order to have better content moderation, it needs to be transparent (so that users can appeal decisions and know why decisions were made), and it needs to contextual, because different communities have different norms that you simply can't automate or generalize. Figuring out good content moderation practices is extremely tough and still an ongoing question, but I want to invest effort into exploring these better practices with Quilt. (Personally, I'm a fan of having visible and paid content moderator users who can help make decisions advocating for community members — somewhat like how subreddits have content moderators, but we'd need to design a process that also prevents them from having too much power over the community.)
• Possibilities for connection, discovery, and organizing. Personally, one of the things that made social media so magical to me growing up was the ability for me to connect with new people around the world. As someone who struggles to socialize in offline communities, I really found solace and connection in online communities, and it feels like so much of this has been lost today. I want Quilt to be a site where we have that ability to connect again (supported by safety practices, transparent algorithmic feeds, and other consentful mechanisms), while also more formally promoting discovery for the many folks who use social media for their work or organizing. I think a lot about what Marisa Duarte has written about how social media has helped to connect Indigenous communities to find each other and explore new communal possibilities. It was also no coincidence that Twitter of the past was a potent place for social organizing, whether from the #BlackLivesMatter movement to #MeToo to so many other instances of making visible local issues, and I do think part of the destruction of Twitter is an intentional act to prevent us from organizing and coming together as community members. I want Quilt to be designed specifically with these possibilities for connection, discoverability, and organizing to be realized, which I think a lot of social media platforms aren't paying attention to much at all today, especially at a time when organizing is so urgent and necessary.
• Creative and customizable. I'm so tired of the boring, templated interfaces that we've been forced to use in a post-Flash, post-responsive design Internet. I miss the Internet of the 2000's where we used CSS to customize our home pages, and I'd like to bring those features of personalized interfaces back to Quilt. I want Quilt to have more personality and be able to reflect each user's uniqueness! Let's bring back personalized home pages filled with images, buttons, colors, and music players.

Finally:

• Cooperative ownership, not federated tech. What ultimately distinguishes Quilt from any other social media platform is the fact that I want to run it as a social media cooperative. For those unfamiliar, a cooperative is an organizational structure that follows the philosophy of "one member, one vote," where each of the contributors to the cooperative are also co-owners. Whereas "federated social media" has taken hold as the dominant response to a post-Twitter social media landscape, I don't think it is the right solution at all. Federated social media is a technical fix to a social/organizational problem, and it's inadequate at addressing issues such as privacy and safety for users. There's been some writing on the shortcomings of federated social media, and I've been wanting to write some more about it myself. But in short, its version of decentralization is inadequate in robustly protecting users from harm and harassment, and the technical commitments of federation force it to abandon certain features such as blocking users and private accounts. While BlueSky has itself attempted to implement some features to address these issues, they're somewhat illusory — blocking doesn't actually work once you deal with multiple instances, and things like user's likes are still entirely public. This is necessary for federated social media to work, and I often regard it as a commitment to the technical protocol over the needs of actual people. Quilt flips this by putting people before the protocol. We'll design the technical aspects of the site based on what features people want and need. Furthermore, I think what will actually help prevent a situation where a company board can unilaterally sell to a rich buyer (Elon Musk), is if we have distributed governance of the organization through the cooperative model. People who use and develop the platform should have a say in how it's grown and managed, and out of all the attributes of Quilt, I find this to be the most interesting, important, and distinct, which other social media sites simply aren't doing.

There are a lot more ideas I have about Quilt, and designing better social media sites in general, but I'll leave it at this for now. You might notice that a lot of my entries link to work from social media researchers, and that's no accident — I truly want this social media platform to be informed by the best research that we have and the actual community members using it. Eventually, I would like Quilt to be a platform where we can support further research into how we can have better, healthier online communities. But with all technical products, we need a minimum viable product (MVP) first, and I think the development of a cooperatively owned, consentful, safe, and joyful social media designed from the margins is already a lot to start with. Is it even possible to build a tech product in this day and age while listening to community members, while protecting their data privacy, while trying to build everything as "ethically" as possible? That to me is what makes Quilt such an interesting experiment.

For those who have followed my writing in the past, you might recognize that my goal has been to develop Quilt, the social media site, through the cooperative known as LUDDENITES, which is a collective whose name plays off of both ludens (or the idea of play as being central to culture and society) and luddites (or the social movement that opposed the use of certain types of automation that led to worker exploitation). Ultimately, I would love to grow the Luddenites collective to build Quilt, but this post is an admission that I simply don't have the expertise, connections, or resources to make it possible. As the idiom goes, "it takes a village," and I really do need a team in order to turn this dream into a reality. And I truly don't know how to find that team (and fund them!). Recently, I've been inspired by the work of Blacksky, this new Kickstarter called Haven Social that's trying to crowdsource something similar to Quilt, and how quickly a developer launched a MySpace-like website called MyOshi for VTubers. While I'm personally not doing a great job of marketing myself and Quilt, I've written this post in hopes that maybe folks will see this and possibly get excited and reach out. Until then, the vision of a consentful, safe, and joyful social media cooperative designed from the margins will have to wait...