PIG: Meet me in the Margins

SoP 2024 RFC
1

Meet Me in the Margins

Request for Comments:
SoP 2024 PIG Grant

Team member names

Nikki Makagiansar & Megumi Tanaka

Short summary of your improvement idea

We seek to improve the way people socialize online by building a browser extension that allows for public commenting on any webpage. Viewers can highlight text and leave contextual comments, start threads, form connections based on interest in subject material that they both happen to be reading.

We believe that deep conversation is more likely to happen when you can start a comment thread on a specific text selection on a website—instead of tweeting a link to an article or a screenshot. You are more likely to meet your community and find people with similar interests when the gathering site is websites you are already visiting whenever you open the web browser.

The concept of user identity will be persistent across every HTTP site you visit, and the comments you leave will be visible to anyone else who is using this extension. In this way, you do not need to sign up for a platform (like X) and try to import your insights to that platform. You can retain the context of a given page by writing notes in the margins.

The key differentiator between our idea and social media platforms is that user data is not mined for advertisement, there is no personalized algorithm, and there is no VC interest in profiting from this idea. It is important to us that this is an open source, not for profit discourse layer that evokes the same feeling as the original websites uploaded to the www in the 90s, before every action was tracked and analyzed. We aim to build tools that are collaborative, interoperable, and open source, rather than scoped to a specific organization.

By adding this layer to the open web, we can gather experimental results to pave the way for further reimagining of what social media means.

meet-me-in-the-margins
meet-me-in-the-margins2458Ă—842 70.4 KB

Q&A

1. What is the existing target protocol you are hoping to improve or enhance? Eg: hand-washing, traffic system, connector standards, carbon trading.

Social media.

2. What is the core idea or insight about potential improvement you want to pursue?

Social media platforms have failed at content moderation. Private corporations like Meta and X have struggled to strike a balance between freedom of speech and trolling behavior. User data is mined to support the interests of the advertising economy, not the interests of the people. Meaningful conversation isn’t the norm on these platforms, it’s the exception.

Instead, the best places to find community are smaller forums such as Discord, Reddit, and private groups where moderators are part of the communities themselves. Instead of competing with this fragmentation by creating yet another platform, we seek to create a layer of community on the surface of the www by creating a social media community that exists in the margins of the sites we’re already on. We don’t want to create yet another app to have conversations with a slightly different algorithm or UI. With our prototype, we will create a way to have conversations on the surface of the open web, without needing to go anywhere else.

3. What is your discovery methodology for investigating the current state of the target protocol? Eg: field observation, expert interviews, historical data analysis, failure event analysis

We follow a standard UX Research framework for all our projects. This means starting with research, competitive analysis, user interviews, and observational studies. Given the timeline, we need to finish researching during the first two weeks so we can spend the bulk of the project period prototyping our software, QA-ing it, and testing with an experimental user group.

Our research methodology will draw upon past and current efforts, goals, and failures of the Web 1.0 and 2.0 era. Many new media writers, technologists, and designers have already laid the groundwork for a multiplayer web, which we will be referencing as it pertains to multiple viewers interacting on a website simultaneously.

We will also analyze failures of anonymous chat systems such as Omegle, ChatRoulette, YikYak, and the Michael’s chat - The Michaels Chat Was Too Perfect For This World - MGRM

We will also research the current state of social media sites on Facebook, Instagram, X, Tiktok, and new products such as Air Chat.

A number of projects for public comments have already been attempted and have failed—we will analyze the failures of Google Sidewiki and Epiverse.co.

For interface design, we will reference contextual annotation on products like Medium, Pocket, Omnivore, and Hypothes.is. We will also reference private sector products like Moonbounce, Superbounce, and Liveblocks.io, as well as Chrome extensions like COLLABORATE and Yawas.

For content moderation, we will research systems of community moderation on Discord, Reddit, and Wikipedia. We are also interested in filters like Gmail spam filters and Akismet.

For commenting systems we will look at HTTP systems like Disqus, Wordpress Comments, and Gravatar. Gravatar is especially interesting as it is an identity system that persists across websites.

We will consider alternative protocols and storage methods such as the fediverse, Gemini, Gopher, p2p systems, blockchain. We also had a wild idea for not using any server space at all, instead building a custom browser that creates a local server SSH tunnel or VPN that connects to other people using the same browser (in this idea, the comments are temporary and only stored on your machine for websites that you have open at the same time someone else is viewing using the same browser).

But most likely we will operate this as a persistent comment system on a standard server setup on the regular http protocol because that’s where most people are.

4. In what form will you prototype your improvement idea? Eg: Code, reference design implementation, draft proposal shared with experts for feedback, A/B test of ideas with a test audience, prototype hardware, etc.

Code. We will design and develop a browser extension (or entire Chromium browser, pending some experimentation) and open it up for beta testing to see what kind of discourse arises.

5. How will you field-test your improvement idea? Eg: run a restricted pilot at an event, simulation, workshop, etc.

We will distribute the software to a group of test subjects who will use it in their everyday web browsing and leave comments for a fixed period. After that period, we will interview subjects and compile our findings on the discourse into a report we will present at an event where we also provide a link to our open source Github repo.

6. Who will be able to judge the quality of your output? Ideally name a few suitable judges.

7. How will you publish and evangelize your improvement idea? Eg: Submit proposal to a standards body, publish open-source code, produce and release a software development kit etc.

We will open-source our software and create a website explaining our methodology and output. We will also market the page within our networks and speak about this project in classes that we teach, panels we speak on, and give workshops within the state of New York to start. We will continue to work on this project beyond the Summer of Protocols.

8. What is the success vision for your idea?

Short term success means encountering a comment written by a stranger on a website we stumble upon on a niche subject. We wish to see this happen personally within 1 month of our event and public beta launch.

Medium term success means getting support in our open source Github repo to improve on the content moderation features and add a collaborative cursor element. We want to hear of stories in our network of people making new connections online. We do not want to see many trolls, and we would like the tool we’ve built to enable users to curate their own layers that show up on each site.

Long term we want to gain funding for this project and form a non-profit. We will continue to promote it across browsers and devices—even developing a browser for iOS and Android. We would like to see different extensions and add-ons built for our service for unique uses cases: such as dictionary definitions for K-12 education and crowd sourced translations for global audiences. These could be URL parameters which filter existing comments in unique, user-defined ways. We would also like to expand upon how people discover content online by creating graph views of how sites are discovered and creating beautiful visuals that represent each person’s unique mode of free association when searching online. Of course, privacy will be the most important value as we continue to build this new kind of social media.

7 Likes
2

Hi @megumi , I’ve been involved in academic research both independently and collaboratively, which drew me to your post.

Academia has an interesting culture, perhaps captured with the commonly used phrase “Publish or perish”. The culture surrounding how knowledge is created and disseminated will make this a very interesting undertaking indeed.

How will your idea support (or run counter to) the often sought after goal of tenure on the part of academics, who themselves are facing the pressure to produce knowledge? This may be a protocol to lay out and further explore in and of itself.

Beyond that, there’s something to be said about culture within academia that merits closer inspection, if you haven’t already done so. Finding ways to co-opt or counteract the “Publish or perish” culture can further pave the way to adoption by your primary audience.

Otherwise, maybe there’s space to shift your primary audience altogether. I feel like the private sector tends to lead the way anyway innovation-wise, with higher education bringing up the rear. Looking forward to your thoughts.

3

Hi @Lenz thank you so much for your question!

The “publish or perish” concept is not one that we are directly aiming to change. I’m not an academic researcher—my project partner and I are both teachers in art schools and alternative schools. Personally I dropped out of UC Berkeley. This is not our domain of expertise so we are not the right people to pursue the tenure track protocol as something to disrupt.

That being said, I am very grateful that you opened up this rabbit hole for me, and I’m having a lot of fun learning more about the replicability crisis. As a future goal for this idea, I’d like to see public annotations on articles that link to p values for reproduced experiments.

I agree with your take on higher education being a difficult primary audience to innovate. There are several examples of collaborative annotation tools within classrooms—Perusall had the best interface of the ones I surveyed—but these tools are behind a paywall and the data remains localized to specific classroom instances.

What we are hoping to accomplish with this project is something broader—commenting on the open web, where no one owns the data. The problems we’ll need to solve are content moderation, online social etiquette, data storage, persistence of comments, user adoption, browser compatibility, device screen sizes, competition with existing comment systems. If we are able to change the way people interact with webpages, I can see it being adopted by teachers much like Google Docs and Slides have been—because they are free.

However, academics are not the primary audience for the first version of the prototype. We were thinking of a closed beta with a series of handpicked independent authors involved in the digital gardening and knowledge management movements—ie., power users of tools like Roam Research, Logseq, and Obsidian. Many of these users are also on the platform are.na and exploring new ways of connecting with people and ideas on the web. We think this community of artist researchers is a great place to start for polite and engaged discourse to set the tone.

I think I’ll have to edit the RFC quite a bit to make this clearer. Perhaps my use of the term “research” is the confusing part—since I meant it in a general sense, not exclusive to academia. Anyone who starts searching for information on a topic is researching.

1 Like
4

@Timber @Lenz @Popline @hpymntn I saw that you all liked this post while it was still called “A Tool for Collaborative Research” ! We have since pivoted and shifted the idea to a more general audience and we are tackling social media now. New project name is “Meet me in the Margins”

2 Likes
5

Love this idea!
How are you all thinking about data retention? Would the margin chat persist? Or would it disappear with some cadence like Snapchat or telegraph? Or would it be set by the community itself ?

6

@Evabungle Thanks for your question!

For privacy, we want to make sure that data cannot be mined or scraped by 3rd parties to form any user profiles, since that will contribute to the ad/attention economy that we’re trying to subvert. For this reason, we are thinking the names of people posting will be kept hidden until a user interacts with the comment by replying. We will keep researching this problem and aim to figure out an interface layer that can be revealed once a user is deemed trustworthy enough, perhaps with a captcha or by agreeing to community guidelines.

As for persistence, the true value of the comments are only apparent once they are persistent, especially for small websites that don’t get a lot of traffic. Larger sites with thousands of active viewers will likely need a temporary commenting mode, or some way to filter out the noise.

We think the critical difference between our idea and dominant social platforms is the ability for the user to control the firehose. Do they want to see everything in chronological order, do they want to see most upvoted persistent comments, do they want to see only those posted by friends or friends of friends, do they want to turn it all off and write private notes just for themselves?

In terms of privacy and persistence, we think there need to be multiple views that are set in the interface—both public/community set rules for the persistent feed, and then a more private or friends only feed. I’m thinking of early Facebook post privacy settings with Public, Friends, Limited, Private.

I do like your idea of a disappearing temporary chat—but I think this is more of a medium term add-on feature that we can consider alongside the multiple cursor view. This is a similar UI to Figma’s cursor chat that disappears after a certain time period. I think our first idea is to launch persistent comments, but an add-on feature would be a temporary margin chat!

1 Like
7

Hi @megumi , I very much appreciate your clarification - very helpful! I’m also in agreement with your recent title change as it helps move the reader away from whatever preconceived notion they might have upon only seeing the word “research.”

While you’re not tackling the “publish or perish” protocol within academia as I had originally assumed, you still certainly have your work cut out for you, in a great way! The primary audience that you described sounds like one with great potential to test-drive this idea. In fact, it seems from what you’ve described that the proverbial pump has already been primed.

Thanks for sharing some of the different tools currently in use, I’m going to check them out. Best of luck with your next steps.