Epistemic Status: Confident about the problem, solution and exact mechanism is more of an untested hypothesis.

There's a saying for communities: if you're not gaining members, you're losing members. Sometimes you hit just the right size and you'd prefer things stay exactly as they are. But in practice, some people will eventually drift away - moving to a new town, getting a new job that consumes more time, or just losing interest in whatever-your-thing is.

So you need to have some source of new members, that supplies the life blood a community needs to keep going.

I think there's a similar truism:

If you are not building organizational capacity, you are burning organizational capacity.

(If the above sounds totally thoroughly reasonable and prompts an obvious course of action you can stop reading here, or skip towards the final section. Most of the rest of the post is for making this more salient)

Young organizations, communities and other groups tend to run on hero power - a couple people who care a lot who put in most of the work to keep it going. This is both because only a few people do care enough to put in that work, and because trying to distribute tasks is often harder than doing the tasks yourself.

As long those organizers are sticking around, everything seems fine. But the system is fragile.

I think people loosely understand the fragility. But I think it's better to think of it as a resource being slowly burned down rather than "a system is working fine until something changes."

Nothing Gold Can Stay - and Gold Melts Suddenly.

At the end of the Lord of the Rings movies, a golden ring is dropped into Mount Doom. When I first watched the film, I expected the gold to slowly melt as it slipped into the magma. Instead, it stayed solidly ring shaped... until it abruptly melted in seconds.

(This was the Ring of Power, approximately 3 seconds before it wasn't anymore)

I thought that looked unrealistic... but someone told me this is how actual gold melts: temperature rising until suddenly it reaches a critical state change. (No comment on magical gold you have to take to a magic volcano)

Any time you have a mission critical job that depends on a person being invested for it to get done, I think you have something similar going on. (i.e. a job where a person isn't getting paid enough for it to be something they sustainably do as part of their general survival, especially if it requires them to exert a lot of agency or continuous attention).

From the outside, it looks like things are fine up until the organizer suddenly burns out. This may even be how it feels to the organizer ("the first stage of burnout is zeal, and the second stage of burnout is burnout"). But under the hood, some combination of thing are often happening:

  1. They're slowly changing as a person, which will eventually make them no-longer-the-sort-of-person-who-wants-to-do-the-thing. They get bored of the thing, or they get interested in new things, or they just need a break.
  2. A slow frustration/resentment eventually builds up that they doing the job without enough help.
  3. They get physically worn out by stresses caused by the thing.

If you notice this far enough in advance, you can train new people to replace you. But this means you need either new heroes willing to do more work than they're paid for, or you need a whole lot of smaller-helpers who are somehow organized enough to do what had previously been a cohesive, unified job.

You may not find those people in time to replace you. Or, they may end up in the same dynamic you were in, and eventually abruptly burn out, and then may not be able to find people to replace them.

Or there may be a multi-step breakdown - you find someone who can do most of the things, but don't quite understand all the pieces, and then when they find a new person to replace them, they find a person who is able to mostly-do the pieces they understand well, but the pieces they understand less well get lost in second step of translation.

Homemade Things Cost More

It costs more to build something yourself than to buy it factory made.

Things you make yourself are often able to be more unique and special than things mass-produced by capitalism. They can cater to special, niche interests without enough demand to develop mass production.

But there was a weird followup step, where Capitalism noticed that people had noticed that homemade things took more time and were worth more. And enterprising entrepreneurs saw free money and learned to market "homemade" things for more money.

As a result, I came to associate "homemade" with "overpriced." Many homemade things aren't that special and unique. An artisinal hand-crafted coffee mug isn't really worth more than a mass produced version on Amazon.

(Maybe this is where Premium Mediocre things come from?)

But... when the homemade thing is unique, when you literally can't get it anywhere else, and you are getting important social or cultural value from it... then... well, if you want that thing, the only way to get it is to pay homemade prices for it.

The problem is you may not be able to pay for them with money. They are usually labors of love. If there was enough demand for them for someone to do them full-time, you'd probably be able to mass produce them more cheaply anyway.

It's unlikely the people making them could actually more easily produce them if they were paid more. Or, the amount of money would be dramatically more than what seems obvious. It's not enough to cover costs. It has to be enough to quit your day job, and then not worry about quitting your day job turning out to be a horrible idea.

This means if you want to pay for a rare, precious thing that you want to keep existing, it is quite likely that the only ways to guarantee it's continued existence is to put in sweat and sacrifice. If things are well organized it shouldn't need to be a major sacrifice, but it may mean serious time and attention that you were spending on other things you cared about too.

I don't mean to say any of this in a moralizing way. This is not an essay about what you "should" do. This is just a description of what is in fact necessary for certain things to happen, if they are things that matter to you.

Solution Hypotheses

My advice on what to do about this isn't really tested, but seem like the obvious things to consider:

For Organizers - Your job is not to Do The Thing. Your job is to make sure The Thing Keeps Getting Done Whether Or Not You Do It.

At first, it may seem nice and high-status to get all the credit for doing the thing. That credit will not sustain you forever - eventually you will probably need help, and it may happen more suddenly than you imagine.

My more speculative conjecture is: as early as possible, no matter what your job is, you should make a part of your job to find new people to start sharing the load.

As early as possible, you should also start investing in systems that make the job easier. Identify wasted motion. And ultimately streamline the onboarding process so new people have an easier time contributing.

This runs against my intuitions because doing the stuff myself is way easier than training a new person to do it. But I think it's important to consider this an essential skill, no matter what your task is. (Even if the primary task isn't very people-centric, you will need to develop the people-skills to identify suitable replacements and train them)

For the People Enjoying The Thing - If you are participating in a thing that is dependent on a few people putting in a herculean amount of effort...

Again, saying this without any intended moralizing, simply as a statement of fact: people can't run on respect and credit forever. And the situation often can't be solved simply by throwing money at it. (Or the amount of money is something like $40,000, to provide enough safety net for a person to quit their day job. Maybe higher if the opportunity cost of their day-job is higher)

If a thing is really important to you, you should consider the amount of effort that's going in, and be aware that this effort is a cost getting paid somewhere in the universe.

Maybe it's worth it to you to put in some portion of the herculean effort.

Maybe it's not - maybe the unfortunate equilibrium really is: "there was one person who was willing to put in 100 hours and a bunch of people who were willing to put in 1 or 2, but not enough to learn the skills necessary for those 100 hours to really work." It may be sad-but-true that it isn't actually worth it to any of the individuals involved to ensure the thing can continue running in it's original form.

But a thing to at least consider is whether, long in advance of the next organizers burning out, you should invest 20 or so hours gaining at least one of the skills that the organizers had developed. So that you can not just chip in with an hour or two of labor, but contribute one of the foundational building-blocks a given event, community or project needed to function.

New Comment
23 comments, sorted by Click to highlight new comments since: Today at 10:13 PM
There's a saying for communities: if you're not gaining members, you're losing members.

This heuristic is totally worth turning into a snowclone and applying almost everywhere. If your net worth is not going up, it is probably going down. If your house isn't being remodeled, it is probably falling into disrepair. If your health isn't getting better, it is probably getting worse. Etc.

The general form of the underlying claim is that that the derivative with respect to time for any measurable characteristic is almost never zero, it is usually either positive or negative, and without attention, the direction is usually not the one that humans typically prefer.

The variant I most often use is: if you're not planning on arriving (to some event or meeting or whatever) early then you're almost certainly going to be late.

It's probably also worth mentioning that achieving a perfect Goldilocks equilibrium, that's stable for more than a brief period, is much harder than sustaining small, but non-zero growth (or just allowing things to fall apart naturally).

Hoping it's okay to drop down to the object level for a bit...something that's been bugging me about events like Solstice is the degree to which they resist being self-documenting. My ideal in this regard would be the Jewish Seder.

If you think about how a Seder works: there is a Hagaddah, where nearly everything that is said or done or sung is listed in the Hagaddah and read from it verbatim - it is an outline, it is a summary, and it is also the primary content. Most importantly, a Hagaddah explains its own structure; it starts out with text that clearly says what we-the-participants are collectively trying to do and why.

And everybody hears that explanation. The structure of the ceremony - what are we doing now, what will we do after that? - is not just intended as private knowledge for people running the ceremony - it's right out there for the participants to know as well. All of them.

So if somebody invites you to a Seder and you walk in blind having never been before, you immediately are given a hook to hang the experience on. "We are here tonight to tell the story of X in the manner Y which we think is important for reasons Z". The person who goes even once is clearly told as part of the event what it was the event was trying to do and if they keep a copy of that Hagaddah on their way out they themselves could run a very similar service next year by making some copies and following the instructions therein.

The fact that Seders are self-documenting and don't change much year-over-year substantially reduces how "heroic" one has to be to run one. You still have to coordinate finding a room and people to fill it and ingredients to put on the seder plate, but once the guests are present running the event itself is easy-peasy; anybody who had ever attended one could pretty easily figure out how to run one.

The polar opposite of that approach would be to invite people to events which appear to be run by a Mysterious Old Wizard, wherein one section is labeled "have somebody get up and deliver a topical original sermon on the theme of being scared and hopeless in the face of certain doom" and another is labeled "have somebody deliver a topical original sermon on the theme of renewed hopefulness in progress and human ingenuity" and another part is labeled "now have somebody perform some original songs they just wrote this month incorporating themes that fit with the aforementioned original sermons". This seems to me like playing the game on Hard Mode. Maybe there's no way around it when you're experimenting with a new format the first time, or even the first few times - the person starting such a thing is a Mysterious Old Wizard who has a wizard's skillset and wants to iterate quickly to see what works - but after you've been at it several years it might be time to lower the difficulty level from Hard to Normal.

Yeah. I actually have a blog post upcoming called something like "The Moment of Darkness Speech may have been my single largest mistake", essentially for the reasons described here.

I have more thoughts on this, essentially boiling down to "By now I do think there is (mostly) enough material for Solstice to switch to Easy Mode, but there are some reasons you want to at least preserve the ability to work in Hard Mode. There are also reasons for somewhere there to exist Big Solstice which has a lot of logistical issues that Small Family Solstice doesn't need to deal with."

More details than that start moving towards "this is basically a conversation for people in the NYC and/or Boston rationality communities to hash things out and make decisions in more detail.

There's an alternate approach I've seen in Neo-Paganism-- have a structure for rituals, and a high proportion of people who can improvise within the framework.

I don't know whether this would work for rationalist rituals (maybe if we start having smaller more frequent rituals), but I'm mentioning it for completeness.

Yeah, I actully think the Pagan approach is almost entirely better, just requiring more skill, and in particular not being able to scale past 30 people or so, which is why I haven't used it for larger tentpole events.


Huh. I don't share the intuition that it can't scale. As long as the improvising ringleaders are on the same page, why would it fail?

I think I basically answered this in the reply to vedrfolnir - more people requires more effort (the simple fact that you need additional ringleaders makes for a harder problem than a single priest scaling their audience). Each ringleader has more work to do the more people there are. Intimacy stops being felt when you don't know everyone there, etc, it's harder to know everyone there if there are more people. Scaling past 30 starts to mean you either need to project your voice loudly or use a microphone, which limits the ability to do soft whispers and/or ruins the pagan aesthetic.

(Also, this is a thing where empirically, it doesn't seem to scale AFAICT. I haven't done a formal study but my impression talking to pagans is that pagan groups rarely do scale past a couple dozen, whereas Christian churches frequently end up in the hundreds, so whether it makes sense or not it appears to be true)


I don't think the barrier to scaling is the practice, but the people. Large group experiences can't be narrowly custom-tailored because their aren't enough people in the target audience; small groups can. People who don't chafe at experiences that aren't narrowly custom-tailored are unlikely to become pagans.

What do you think the barriers to scale are? Just difficulty in finding people comfortable with improvising within the framework to that many people, or?

Among other things, yeah - it requires more effort on the part of everyone. And the effort required scales with the number of people involved. The more people who are improvising, the more people you have to be paying attention to and modeling at once, the more you have to trust them, etc.

Also, the sheer trust and intimacy is an important part of it, regardless of complexity

I think this comment is actually a pretty important piece of "how to do Solstice" that should be entering into the "How to do Solstice" canon. It also makes me feel a bit better about including both 

a) a more extensive Welcome Speech, and

b) the Meditations on Past/Present/Future in the 2020 Solstice despite those not being that popular, since I think they provide important framing for the whole thing that makes it more sticky.

(I don't know that they actually solve the problem either, but, I think things in this space are important for Solstice)

Your solutions remind me a lot of what we in the business world think of as protecting the business against existential risk from the failure of individual human resources, cutely measured as a "bus factor" for the number of people who have to get hit by a bus before your business is in trouble. A great source of information about techniques for dealing with this, at least within a small business context, has for me been The E-Myth by Gerber, but you can also look at how continuity planning is done (although there the focus is usually on seemingly exceptional rather than normal events (though of course the exceptional is normal, and that's the whole point of doing continuity planning)).


The "bus factor" you mention reminds me of "The Secret Life of Passwords" , a NYT article that discusses, among other things, how a financial services firm went about trying to guess business critical passwords after most of their employees were killed in the 9/11 attack.

Howard Lutnick, the chief executive of Cantor Fitzgerald, one of the world’s largest financial-services firms, still cries when he talks about it. Not long after the planes struck the twin towers, killing 658 of his co-workers and friends, including his brother, one of the first things on Lutnick’s mind was passwords. This may seem callous, but it was not.
Like virtually everyone else caught up in the events that day, Lutnick, who had taken the morning off to escort his son, Kyle, to his first day of kindergarten, was in shock. But he was also the one person most responsible for ensuring the viability of his company. The biggest threat to that survival became apparent almost immediately: No one knew the passwords for hundreds of accounts and files that were needed to get back online in time for the reopening of the bond markets. Cantor Fitzgerald did have extensive contingency plans in place, including a requirement that all employees tell their work passwords to four nearby colleagues. But now a large majority of the firm’s 960 New York employees were dead.

I don't think it is realistic to aim for no relevant knowledge getting lost even if your company loses half of its employees in one day. A bus factor of five is already shockingly competent when compared to any company I have ever worked for, going for a bus factor of 658 is just madness.

Just to chime in with support, I read "The E-Myth Revisited: Why Most Small Businesses Don't Work and What to Do About It". It is not obviously epistemically sound because it is offered more like "lore" than "science". However it stayed with me, and seems to have changed how I approach organizational development, and I think I endorse the changed perspective.

One of the major concept handles that may have been coined in the book (or borrowed into the book thereby spreading it mnuch further and faster?) is the distinction between "working in your business versus working on your business". A lot of people seem to only "work in", not "work on", and book makes the claim that this lack of going meta on the business often leads to burnout and business failure.

One thing to keep in mind is that since all debates are bravery debates and this specific community is often great at meta, it is also possible to make the opposite error... you can spend too much time working "on" an organization, and not enough "in" the organization, and the failures there look different. One of my heuristics for noticing if there is "too much organizational meta" is if the bathrooms aren't clean.

One of my heuristics for noticing if there is "too much organizational meta" is if the bathrooms aren't clean.

Heh, that feels right to me. Although I'd phrase that not necessarily as "evidence there is too much meta" so much as "evidence there is not enough object level." (which may or may not point to the former, depending on context)

Another vote for E-Myth here. I somehow managed to pick it randomly off a bookstore shelf many years ago, long before I had ever done anything entrepreneurial myself (I was in HS or college at the time.) But it really stuck with me, and I still have my copy.

I think it's worth thinking about how much to turn the thing into a machine that keeps the events and meetings churning as they have been on easy mode, and how much to have new heroes that turn the organization into a new labor of love that is in a slightly different direction. In one direction, you're building systems, and in another direction you're building autonomy and instigatingyness.

I think there can be a problem where when people are working in a system and it is easy and pretty good, they aren't quite paying attention to the things they would really want if they were running things --- and often those ideas are better than the system. The simplest way that this occurs is when people do the minimum the system asks, vs when they are working with the system and using it to help them make things awesome.

People can be influenced to move towards the self organizing direction by asking them what they think would be awesome, and getting details of what's cool about that and why they want, and providing social capital / organizational capital etc to help implement their thing. It's usually well received if you are trying to help them get more of the thing that they want. There might be some other things that are important in this process, I'm not sure what those are.

In this vein, I would be very interested in hearing anecdotes about how easy mode events feel different from hard mode events. I don't think I've ever participated in an easy mode event that did not feel like a poor use of time, but that might be due to the environments where those happened (schools and universities).

So the way I'm _hoping_ to think of this is in having a machine that outputs agency. In particular, that's what I think most of the point of having meetups is.

Specifically in a rationality-community-paradigm, I think a good framework is something like "have a roster of fairly easy-mode events you can run for low effort, but as often as possible try to have an effort that requires something to put something interesting together."

I think having a basic framework of "people take turns making sure at least something happens each week", and anyone who puts effort into a cool/different thing happening will tend to get bonus status, and people can try out new bigger or wildly-different things on non-meetup-days if they want.

as early as possible, no matter what your job is, you should make a part of your job to find new people to start sharing the load

But what if you're irrayplaceable?