The website you're currently using is powered by ForumMagnum. But what really is ForumMagnum? What is it for, and why was it designed this way? In this post, I cast ForumMagnum as a medium for building communities of inquiry. I show how ForumMagnum is designed to build norms like rationality and long-form. Lastly, I suggest how the ForumMagnum developers could use the body of CoI research to guide their future product design.

What is ForumMagnum for?

ForumMagnum describes itself as "the codebase powering LessWrong and the Effective Altruism Forum."^[1] That's the what ... but what's the why? Here's why I believe ForumMagnum exists:

ForumMagnum is a medium for building online communities of inquiry. It's a web forum that embeds norms of rationality and long-form comms.

That is: ForumMagnum is not defined by its users, or its features, or its codebase! ForumMagnum is a medium designed to carry a message, and its message is a set of social norms. Let's see what those norms are, and how ForumMagnum's features are designed to build those social norms.

What are Communities of Inquiry?

In the most cited paper of all time, Garrison et al. set out to "investigate the features of the written language used in computer conferences [e.g., forums] that seem to promote the achievement of critical thinking."^[2] They built on the concept of a Community of Inquiry (CoI), which Wikipedia nicely defines:

The community of inquiry (CoI) is a concept first introduced by early pragmatist philosophers C. S. Peirce and John Dewey, concerning the nature of knowledge formation and the process of scientific inquiry. The community of inquiry is broadly defined as any group of individuals involved in a process of empirical or conceptual inquiry into problematic situations. This concept was novel in its emphasis on the social quality and contingency of knowledge formation in the sciences ...^[3]

The LessWrong community, the EA community, the Alignment community, the Progress community — what unites them? It's not just that each community uses ForumMagnum! It's that each community is a CoI. Let's see how each community describes itself.

LessWrong is online forum/community that was founded with the purpose of perfecting the art of human rationality. ... LessWrong is a good place for [those] who want to work collaboratively with others to figure out what's true.^[4]

Effective altruism is ... both a research field, which aims to identify the world’s most pressing problems and the best solutions to them, and a practical community that aims to use those findings to do good.^[5]
The Alignment Forum is a single online hub for researchers to discuss all ideas related to ensuring that transformatively powerful AIs are aligned with human values.^[6]
The Progress Forum is ... a place for long-form discussion of progress studies and the philosophy of progress. ... The broader goal is to share ideas, strengthen them through discussion and comment, and over the long term, to build up a body of thought that constitutes a new philosophy of progress ...^[7]

None of these communities explicitly identifies as a Community of Inquiry. Nevertheless, each community neatly fits the definition (at least, aspirationally!). Each community emphasizes its own problematic situation, to be solved with inquiry, and which is best done socially.

How ForumMagnum builds Communities of Inquiry

As such, the ForumMagnum communities share similar norms.^[8] They're nicely summarized below each comment box:

Aim to explain, not persuade
Try to offer concrete models and predictions
If you disagree, try getting curious about what your partner is thinking
Don’t be afraid to say ‘oops’ and change your mind

But how does ForumMagnum build these norms? There's one norm-building technique everyone's aware of: moderation, or manual brute-force. Since this technique is very well-known, in this post, I'll focus on subtler techniques.

Technique 1: State your norms

We actually just met the simplest norm-building technique. I call it state-your-norms. The norms are written below the comment box, shown at exactly the right time. I believe this technique is extremely powerful. I'm willing to bet that it's as influential as all manual moderation.

Yet it's surprisingly under-used! I bet your enterprise Slack or Notion has no such norms stated anywhere, let alone next to the comment boxes.

Technique 2: Friction

Speaking of that comment box: consider the behavior of the ⏎Enter button there. There are two viable behaviors ⏎Enter could have: either to post your comment, or to create a new paragraph. In ForumMagnum, ⏎Enter creates a new paragraph, and there is no keyboard shortcut for “Submit”. This is deliberate.

If the ⏎Enter button sends my message, I’ll write short messages. But if creates a new paragraph, I’ll write longer comments. This product design builds the norm of long-form, async communication. This is an important norm on these sites, although not usually made explicit.

This is an example of a norm-building technique that I call friction. The word “friction” in UX design is often used negatively, but friction is a powerful way to steer users towards desired behavior! ForumMagnum uses several frictions to build the long-form, async norm. Notice there are no realtime notifications, and timestamps are only accurate to the hour.

Technique 3: Controlled feedback loops

An aside: I bet you’ve experienced what I call Shift-Enter anxiety. Visiting a new app, will the ⏎Enter key create a new paragraph, or will it prematurely send my message? With this micro-stress, I must make a guess: does it look like this app wants long-form writing? Is the input box large? Are other people posting multiple paragraphs? If so, I hit ⏎Enter, and pray for a new paragraph.

That example shows norming feedback loops at play. After I post my longer comment, I help build the norm that this is a place for long-form. Nudge theory told us that defaults are powerful. But in multi-user platforms, defaults are all-powerful. As @Viliam wrote, you become the UI you use.

Media designers can strengthen feedback loops with voting. This lets users reinforce norms. But it's risky: if a bad behavior becomes a norm, it will also be reinforced by voting systems!

One way ForumMagnum guards against runaway feedback loops is with named reactions. In most apps, you can vote/react with emojis. But emojis can be very ambiguous (even causing legal issues!). In contrast, ForumMagnum's reactions have labels, like: “Changed My Mind”, “Insightful”, or “Good Facilitation”. This incorporates the state-your-norms technique, by embedding the norms in the reactions. It also incorporates the friction technique, by making it harder to react in undesired ways (for example, there's no "Too Long, Didn't Read" reaction).

Technique 4: Zoning

I'll end with a norm-building technique that I call zoning. Counter-intuitively, one way to build a norm is to build a feature for its opposite norm! It's like how urban planners try to move the pollution to an industrial zone, or the prostitution to a red-light district. LessWrong has two such zoning features.

One zoning feature is called Shortform. “Exploratory, draft-stage, rough, and rambly thoughts are all welcome on Shortform.” Implicit in this description is that such content is not generally welcome elsewhere. The “shortform” feature says: “Normal posts are long-form and carefully edited.”

Another zoning feature is called agreement voting. So many other forums are plagued by group-think. It's a malignant norm. ForumMagnum attempts to zone the groupthink by moving it to a separate voting axis. The normal voting axis is the traditional “How much do you like this?”. The second axis is: “How much do you agree with this, separate from whether you think it’s a good comment?”. The “agreement voting” feature says: “Normal voting should not consider agreement.”

How ForumMagnum can use CoI research

Anyway, back to Garrison et al., and their community's 25-year-long investigation into how text-based media can promote rationality. How can CoI help ForumMagnum and its communities?

The CoI community provides a framework for defining inquiry. It has some great insights. For example, that social presence (the ability to present yourself as a "real person") is essential to inquiry. I suspect that the lack of social presence was one reason for the failure of Arbital (a sort of wiki-structured LessWrong).

The CoI community provides a survey for measuring inquiry, and attempts to measure inquiry using AI. The idea of "measuring intangibles" is central to the ForumMagnum communities. Despite this, each community defines its aspirational norms, but has never measured whether they are in fact norms.

The CoI community provides empirical research into how to build inquiry. For the practically minded, try Holly Fiock's big list of concrete recommendations!

How is ForumMagnum really developed?

I described all this as an outsider looking in. How do the ForumMagnum developers really approach product design? I don't know! I believe most development is done by the Centre for Effective Altruism. If any of the developers are around, I'd love to hear their view!

^{^}
Quoted from the ForumMagnum README.
^{^}
Garrison et al., "Critical Inquiry in a Text-Based Environment". Found at position 6884 on Lens.org's list of works sorted by citation count.
^{^}
Wikipedia, "Community of Inquiry"
^{^}
LessWrong, "New User's Guide"
^{^}
EA Forum, "What is effective altruism?"
^{^}
AI Alignment Forum, "Welcome & FAQ"
^{^}
Progress Forum, "About us and FAQ"
^{^}
For more detail, see an unofficial LessWrong norm list, the official EA norms, and the Progress Forum participation guide. I think if you extract the commonalities, you get something like the guidelines below the comment box.