TL;DR

I spent eight months as a community builder for an EA group. This is a reflection on what that work actually looks from the inside, why I’ve found it’s structurally undersupported, and what I think could be different for funders, organisations, and for community builders themselves.

What community builders actually do

From the outside, community building can look deceptively simple. Organising events, running fellowships, hosting discussions. And yes, those things are part of it. But most of the actual work happens in the spaces between. The one-on-one conversation with someone trying to figure out how they want to contribute. The introduction you make between two people who probably should've met a year ago. The moment someone finally feels comfortable asking the question they've been sitting with for months because the room feels safe enough.

These things don't show up cleanly in metrics. But they tend to be the moments that actually shape people's trajectories.

If you think about how someone typically moves toward an impact-focused career or project, it rarely happens in one step. There's an initial spark, be it through an event, an article, or even a conversation. Then, there’s deeper engagement. After that, usually what comes are relationships with people asking similar questions. Then, slowly, a sense that doing something meaningful is actually possible. And eventually, a decision.

Community builders mostly live in the middle of that process. We're not producing the final outcomes directly. But we help create the conditions that make them more likely to happen. That's the work. And it's a lot harder to see than a headcount at an event.

The call that made me write this

I spent more hours than I'd like to admit on that report. Not because anyone asked me to (they actually recommend not to spend anything more than 3 hours doing it) but because I wanted whoever read it to actually understand what we'd been building. The quiet trajectories. The person who showed up to one event and two years later had mapped out a new career, the city slowly becoming a hub for Spanish-speaking people trying to figure out how to have a positive impact.

The call went fine. Friendly, even. We answered the questions. We explained the work. But at some point, almost as a side note, the person evaluating us mentioned that the way they assess community building grants probably isn't the right approach. They were being honest, and I feel that even reflective. I appreciated that.

I just couldn't stop thinking: we spent months trying to meet a bar that the person holding it isn't confident about either. And please believe me when I say this: it’s not a criticism of them. It's the thing I keep coming back to when I think about what community building in EA actually is, and why supporting it is so hard to get right.

The strange place community building holds

In a short span of time, I’ve got to see that community building in EA holds a somewhat strange place. It's not exactly dismissed; nobody says it doesn't matter. But at times it feels like there's a subtle hierarchy at play, the kind that doesn't “announce” itself. Research gets funded. Direct work gets funded. AI Safety, well, that's a different conversation entirely. The people making sure there's a room full of engaged humans willing to do that research and that direct work? Well, that's where things start getting complicated.^[1]

I got to see it from the inside. There are programs designed to support community builders. Resources, frameworks, training. And they're very useful, up to a point. The problem is that information without infrastructure tends to feel just like homework. You can know exactly what a thriving community looks like and still have no clear path on how to build one, because the structural support (the runway, the feedback, the stability) isn't quite there.

What you end up with is people running a professional operation on hobbyist fuel.^[2]  Passionate, capable people doing serious work, held together by peer support and goodwill, in a system that hasn't quite decided whether they're worth investing in properly.

What doesn't show up in a grant application

That person you've been quietly supporting for weeks, or even months, as they figure out where they fit in this space. And not because you want to show some numbers in a report, but because you can see their potential and they just need someone to think out loud with. That initiative that exists because you made two people meet at the right moment, or the message you get six months after an event from someone saying they finally made the career change they'd been afraid to make.

These things happen. I watched them happen. And every time they did, I'd think: how do I put this in a spreadsheet?

The honest answer is you can't. At least not properly, that I know of. You can count events, attendance, pledges, career shifts (and you definitely should, because those things matter!) but they're the visible tip of something much larger. A community isn't just the sum of its measurable outputs. It's also the environment that makes those outputs possible. The belonging that keeps people engaged long enough to take the next step. The trust that makes someone willing to change their life based on a conversation (or a couple).

Someone in the community pushed back on this recently. After we published a post about EA Barcelona's transition, we received some internal feedback, asking whether we had any internal model for measuring the marginal impact of community building. I don’t think they were being dismissive, but rather that it was an honest question. They used a janitor analogy to relate it to a community builder's work, saying that a janitor prevents deterioration, which is measurable by its absence. I think it's a fair point. I didn't have a clear answer then, and I still don't.

But I believe community builders are something different. We're not just there to prevent deterioration. We're there to create the conditions for things to exist that otherwise wouldn't, or that would at least take much longer. I know that's probably harder to see, but it doesn't mean it isn't there.

The person who showed up quietly to an AI governance talk, and two years later had narrowed her career to three high-impact paths she hadn't known existed. The Erasmus student who attended events for a semester, went home, and later influenced his employer to donate €12,000 to effective charities. The person who first came in unsure how to contribute and ended up co-leading the first national retreat. None of that shows up cleanly in a report. But it happened. I was there.

What’s actually broken, and what could be different

I want to be honest about something before going further. Funders have real constraints and genuinely hard tradeoffs. The argument for not funding every community isn’t “community building doesn’t matter”, it’s “we can only fund the highest-impact work and we’re trying to identify what that is”. I understand that, and I respect it.

But here’s what I keep coming back to: are current evaluation methods actually identifying the highest impact communities? Or are they identifying the most legible ones; the ones whose impact is easiest to see, count, and put in a report? The communities that are easiest to evaluate are often the most recently funded, the most structured, and the most willing to optimize for reportable outputs. That's not the same as the most impactful. And if we're optimising for legibility instead of impact, we might be systematically cutting exactly the kind of slow, relational, hard-to-measure work that makes communities actually function.^[3]

EA values epistemic rigor, and rightly so. But community building runs on social processes that resist that kind of rigor by design. Trying to evaluate one with the tools built for the other is part of why this keeps going wrong.^[4]

I've been thinking about this a lot, and I don't think it's any one person's fault. I think it's most probably a design issue.

The first thing is feedback. And I’m not talking about the kind of feedback that comes after a decision has already been made, but the kind that helps you understand, while you're in it, what's working and what isn't. At some point we explicitly asked our funders: "what would make this work stronger? What are we missing?" And the answer we got was vague enough that we ended up in exactly the same place we started, and honestly, I'm not even sure the system had a clear answer to give. In the end, what would've made the biggest difference wasn't more resources or more training. It would've been something clear to reflect on: here's what a thriving community looks like at your stage, here's where you're doing well, here's where there's room to grow. That kind of feedback that helps the CB in front of you, and it becomes knowledge that travels to other CBs, other cities, and other communities trying to figure out the same things in different languages and contexts.

The second thing is something I wish someone had told me on day one: don't wait until your community is "mature" to think about diversifying your funding. The model of securing a grant, building, and then figuring out sustainability later puts you in an incredibly fragile position. You're always one decision away from everything stopping. If I were starting again, I'd be looking for alternative funding sources from the very beginning. And not as a plan B, I’d be looking for them as part of the main plan.

The third thing is harder to name but probably the most important. Community building needs to be treated as infrastructure, not as a nice-to-have. Every big idea that has actually changed something in the world didn't just happen because the idea was good. It happened because people carried it forward, argued about it over dinner, convinced their friends, and built the rooms where the right conversations could take place. The Montgomery Bus Boycott wasn't spontaneous. It was years of organising and groundwork laid by people whose names most of us don't know. The plant-based food movement didn't go mainstream on the strength of the argument alone. Someone built the rooms. Someone made the introductions. Someone showed up consistently enough that other people felt it was worth showing up too. That's what community builders do. 

Until the systems that fund and support this work actually reflect that, not just in words or funding, but also in how they give feedback, how they set expectations, and how they create runway, we'll probably keep running professional engines on hobbyist fuel.

Communities aren’t background noise

I didn't write this to settle a score. I wrote it because I spent eight months watching something real get built, and I think the people who genuinely care about impact should know what it takes to build it.

Communities aren't background noise. They're where ideas become movements. Where someone who's been thinking alone for years finally finds a room full of people asking the same questions. Where the next person who changes something important first figures out that changing something is possible.

If you're part of a community that's shaped how you think or what you do, consider that somebody made that possible. And consider what it would mean to treat communities not as a byproduct of the work, but as the work itself.

And if you’re a CB reading this: I don’t think we’re passive victims of a broken system. We have more agency than we sometimes give ourselves credit for. Ask for feedback explicitly and early, not just at grant renewal time. Start thinking about funding diversification on day one, not when you're already fragile. Document the invisible work, not because funders will always value it, but because it helps you see what you're actually building. The system needs to change. I'm not sure it will. But I think naming it clearly is at least a start.

1. ^{^}

    To be fair, structures do exist. The Community Building Grants (CBG) program, the Organiser Support Program, Groups Support Funding… these are real attempts to address this, and I'm really grateful they exist. But there's a difference between funding that acknowledges community building matters and funding that's designed to help it thrive. What I experienced, and what I've heard from others in similar positions, is mostly the former.

2. ^{^}

    By "hobbyist fuel" I mean the kind of energy and resources that sustain something people do out of passion in their spare time, peer encouragement, personal motivation, or even volunteer goodwill. These things are really valuable, but they have limits. They're not designed to sustain full-time, professional-grade work over the long term. When community building relies primarily on these resources instead of stable institutional support, it works… until it doesn't. And when it stops, it usually stops abruptly.

3. ^{^}

    While researching for this piece I came across a 2019 EA Forum post raising almost exactly this concern that funding decisions for community builders worldwide were concentrated in very few hands, with no comparable diversity of funders to what exists for other cause areas like AI safety research. The post has since been deleted, but an archived version is still accessible. What struck me was the argument and, believe it or not, the date. This was identified six years ago. From my short-lived experience, the structure hasn't fundamentally changed.

^{^}

 A recent discussion on the EA Forum about community building raised a similar point: that we may be optimising for metrics that fit neatly in reports rather than those that actually predict long-term impact. See: Sam’s Hot(?) Takes on EA Community Building (September 2025).

Show all footnotes