Rationalist Community Hub in Moscow: 3 Years Retrospective

berekuk25 Aug 2018 0:28 UTC
171 points
23 commentsLW link

Short summary: Moscow rationalist community is organized around a local venue called Kocherga. Kocherga hosts 40–50 rationality-adjacent events per month, teaches a few dozen people per year in CFAR-style applied rationality workshops, and has ambitious plans for further growth.

We on the verge of becoming profitable, but we need some funding to stay afloat. We’re launching a Patreon page today.

This post is a long overdue report on the state of LessWrong Russia community and specifically on the Kocherga community space in Moscow. This post is also a call for donations.

Russian LessWrong community has existed for many years now, but there wasn’t much info on lesswrong.com about how we’re doing. The last general report by Yuu was posted in March 2013. Alexander230 made a few posts about Fallacymania which is one product, but it’s far from the only interesting thing about LW Russia.

In the following text I’m going to cover:

  • History of LW Russia from 2013 to 2015

  • History of Kocherga anticafe, the rationalist community space we started in 2015

  • The list of Kocherga’s successes and failures, as well as some tentative comparisons between Kocherga and Berkeley REACH.

  • Our current financial situation, which is the main reason I’m taking the time to do this write-up now.

Current state of Russian and Moscow community

Here’s a short description of LW Russia and LW Moscow in its current state.

Our Russian Slack chat has 1500+ registered members, 150 weekly active members and 50 weekly posting members.

LessWrong.ru has a few hundred posts from Sequences and other sources (e.g. from SlateStarCodex) translated by volunteers. LessWrong.ru gets 800 daily unique visitors. There’s also a wiki with >100 articles and a VK.com page with 13000 followers.

There are regular meetups in Saint-Petersburg (hard to measure, they had a few pivots in their approach in the last few years) and in Yekaterinburg.

Our Moscow community hub, Kocherga, which is the main topic of this post, hosts 40–50 public events per month (45 events in June). The attendance varies from 3-5 visitors for smaller events to 20–30 visitors for larger ones.

We have ties with a local skeptic society, critical thinking crowd, pop-science communicators, a local systems management school and a local transhumanist community.

In the following text I’m going to talk mostly about LW Moscow and Kocherga, because that’s what I’m involved with and that’s what occupies most of my time. I’ve been working on Kocherga full-time since 2015, funding it by myself, and I’d love to be able to keep doing it indefinitely.

Some more specific stats about Kocherga:

  • We get 1000 non-unique visits per month.

  • The lower bound on the number of unique visitors per month is 150, but that doesn’t include “anonymous” visitors who don’t have a club card and so can’t be tracked across different days; also, it doesn’t say anything about the number of rationality-adjacent visitors (some people use Kocherga space for their own goals—coworking, etc.)

  • The total number of people who participated in our paid applied rationality courses and workshops over the last 3 years is 150+, not including a few corporate trainings and summer camps.

  • Our monthly revenue is $6000–$9000, depending on the time of the year and on whether we skipped a workshop that month.

  • Our monthly expenses is around $8500. (Half of it is rent, I’ll provide a detailed financial breakdown below.)

  • We have maaaybe 2–3 months of cash runway left at this point :(

  • Almost all our events are either directly connected to rationality (LW meetups, rationality dojos, nonviolent communication practice, Sequences discussion) or at least rationality-aligned (topics vary from pop-sci to board games).

LW Russia before 2015

Here’s a very short history of LW Russia before Kocherga, from my perspective:

  • There was some activity in 2011–2012 before I joined. (Yuu and turchin were among people who organized those meetups.)

  • I joined in January 2013; I was working at Yandex at that time and was able to provide a space for meetups in their office.

  • 2013 and 2014 were a period of a big growth: we started with 8–10 attendees per meetup and grew up to 30–40 and sometimes 50 attendees by mid-2014.

  • By the end of 2014 we had some drama and a setback due to some of the key members leaving.

  • End of 2014 was the first time we made a serious attempt to put together a CFAR-like workshop, but that first attempt didn’t succeed because of the aforementioned drama.

  • At the same time Pion and I have decided that opening a rationality hub would be awesome and started scheming.

  • We never got back to having 50 concurrent attendees (our space is not big enough to fit so many people anyway), but I feel like the community is significantly larger now than it was in 2014. We compensate for the lower per-event attendance with the total number of events.

Activities

Content-wise we had:

  • Regular bi-weekly meetups with talks on various topics:

    • lightning talks on fallacies and biases

    • various rationality-adjacent topics

    • random sample from one of the events: symbolic self-completion theory talk; binaural rhythms talk; training game; how to measure anything talk; debates in the “Karl Popper” format

  • A few people (Alexander230 and others) developed the Fallacymania game and started playing it regularly.

  • We began to do video recordings.

Internal currency

We tried to organize a prediction market a few times but people lost interest quickly. We still wanted to get people to bet on their beliefs, though, so we had an idea of an internal currency which people could use to make one-on-one bets, without going through the bottleneck of a shared whiteboard. We called this currency “yudcoins”. No, we’re not a phyg, I promise.

At first, yudcoins were emitted on each meetup separately. We chose somethings new and cheap every time — pasta tubes, buttons, rubber bands, etc. Later, when we opened Kocherga club, we printed cards.

Kocherga anticafe (2015—…)

We got all preparations in order and found a relatively affordable venue to rent by August 2015 and opened it to the general public in September 2015.

I covered the initial spendings and most of the following losses from my savings. (I had a stock option bonus from Yandex, about $50k-100k in total, and that money was the reason Kocherga could exist for the last 3 years.)

By the way, the name “Kocherga” (“Кочерга”) means “fireplace poker” and it’s a reference to the Wittgenstein’s Poker incident.

Location

We’re located in a semi-basement with 4 rooms:

  • The Lecture Room (50 sq.m., the largest one; it fits 20 people comfortably and 30-40 people whenever necessary)

  • The GEB Room (30 sq.m., walls decorated with the Godel’s portait, Escher paining and Bach’s music)

  • The Chinese Room (17 sq.m.; it had a chinese-styled wallpaper when we moved in and we couldn’t pass the opportunity)

  • The Summer Room (9 sq.m., the smallest one; we installed metal-halide lamps in there so it can be used for the light therapy; Moscow winters are cold and tough)

You can find the photos of all our rooms here.

Anticafe

Kocherga is an anti-café (Wikipedia uses an “anti-café” spelling, but i’ll use “anticafe” from now on because “anti-café” is too annoying to type), which is a popular business model in Russia: we charge a small fee for the time spent and everything else is free (tea, coffee, cookies, board games, rationality events). The fee is 2.5 rubles/​minute = $0.04/​minute. There’s also an option to buy a monthly pass for $40–$80.

A stereotypical anticafe is targeted at college and high school students. Anticafes are places to hang out and play board games and video games. Here’s what’s on a schedule of events for a typical anticafe: board games; checkers; Mafia; checkers again; movie club; poetry night; board games again.

So we try to meaningfully differentiate ourselves from other anticafes.

  • We have a shelf with board games but don’t host generic board games events like other anticafes do.

  • We avoided video game consoles for a long time; we have two PlayStation consoles now, though.

  • We’re careful with what we put on our schedule, in general: no art classes, poetry evenings or yoga trainings (we could reconsider if the event is organized by the known community member, but we won’t accept such events by outsiders in order to establish a Schelling fence).

For example, here’s our schedule for a typical week:

Monday:

  • Rationality dojo on premortems.

  • Math seminar on combinatorics.

Tuesday:

  • Street epistemology practice.

Wednesday:

  • A talk by Viktoria Valikova from the Health&Help project.

  • Game night: Fallacymania and a few other games created by Alexander230.

Thursday:

  • Nonviolent communication practice.

Friday:

  • Sequences reading club.

Saturday:

  • A talk on the topic of cognitive reappraisal.

Sunday:

  • regular LessWrong meetup; topics covered in the meetup included:

    • CFAR rationality habits checklist;

    • Great Depression and Keynesian economics;

    • EDT vs CDT and other decision theories.

Well… I hope that by this point in my post I provided enough evidence that we’re a legitimate and at least kinda successful rationalist community and not some random people who are trying to get our hands on easy money. (Because getting enough donations to keep our project alive is definitely the goal of this post.)

Kocherga vs Berkeley REACH

The comparison with Berkeley REACH is unavoidable at this point. It’s hard to make a fair comparison since I’ve never visited neither REACH nor the Bay Area, but I’ve read Reflections on Berkeley REACH and I can notice the following key differences:

  1. We charge money for visiting. This definitely cuts off some potential members, but it also allows us to pay (some part of) the rent, so we can stay open to the general public and provide the refreshments. (I’d expect the space to become flooded by the general audience looking for a coworking space if the visiting was no-strings-attached free, and then we’d have to look for a different barrier to entry.)

  2. We don’t provide a place for the people to crash. There are several reasons for that. First, while we have foldable couches, we don’t have a shower or a real kitchen. Second, we’re located in a basement of an apartment house, and its residents could decide that we’re an illegal hostel and probably get us evicted if they disliked us enough.

  3. We have a team of 6–7 paid receptionists who meet the visitors, track their time, keep the space clean, answer the phone, etc. Not all of them are committed rationalists, but most are at least partially interested and excited about the topic.

We also have visitors tracking solved (the solution is standard for all anticafes): each visitor opens an order when they come in and close an order when they leave. So we get plenty of data on whether we’re growing, what’s the average visit duration, etc.

Successes

We worked through a huge amount of details in the last three years, and it’s hard to compress it to a short list of achievements. Logistics, HR, upkeep, IT, design, advertising, knowledge organization — each of these areas took plenty of thinking and doing to work through. Of course, each of these areas is always a work-in-progress.

But the key points which I’d consider to be visible outcomes are these:

  • We kept our focus on rationality and have a reputation of being the place for everything rationality related.

  • We have a healthy community with a high bus factor.

  • We have a program for our two day rationality workshop (and sometimes three-week course with the same content but the different pedagogical approach) which we held for 15 times since 2016, totaling 150+ graduates.

  • We built the infrastructure (both technical and organizational) for managing our day-to-day operations.

Focus and reputation

From the very beginning, I believed that we should be a community with a small cult-like following, tightly focused on the core rationality topics. Over the years I have slightly updated in the direction of being a little bit more inclusive, but I still believe we should be wary not to lose our core values along the way (although this update feels more like understanding our boundaries in more details and not like a single “how radical vs inclusive are we?” number).

This belief caused a lot of disagreements. I had the following conversation dozens of times:

Somebody: “You should have more events on a wider amount of topics — art, history, public speaking, music jams. You’ll attract more people if you do this, I’d hate to see Kocherga go under because its areas of interest are so obscure. Also, I don’t understand math so I feel excluded from your rationality discussions.”
Me: “We need to have some kind of a boundary on which events we’re happy to host and which events we try to avoid. The topics you’ve mentioned are more popular, which means that we’d risk drifting into the typical mainstream point in the anticafe-space. That would be bad because we want to attract disproportionally more rationality-minded people by signalling that our interests are not typical. I’d rather have a Shelling fence and lose some amount of potential visitors and income.”

Still, while we tried to keep our focus, we successfully integrated into the large pop-science community:

I can’t give the hard numbers, but it feels like we’re well-known, respected, and don’t have a reputation of being a crazy cult. I periodically stumble upon comments and posts from people who don’t live in Moscow which reflect their disappointment that they can’t participate in our community. We raised $7000 last year in a crowdfunding campaign and the response was very positive.

On the negative side, our core members are, on average, more neurodivergent than the general public (with a large individual variation), as you can expect from any rationalist community, and to those who consider high Guess Culture and social awareness skills to be a basic requirement this can be aversive. I’d love to raise the average social skills level in our community, but in the meantime I’d still take a math-savvy but not-well-adjusted-socially person over the opposite any day of the week.

Community and events

We’ve scaled the number of events from zero to 40–50 per month.

Here’s how we got here:

  • Our Sunday LW meetups are organized by several different people and are currently at the point where I don’t have to spend any effort to make sure they’re happening.

  • Some regular events got launched without my participation by the community members (e.g., Street Epistemology).

  • Some regular events were bootstrapped by me or someone else and then passed over to the new organizers.

    • e.g., I lead our rationality dojos for the first two years, and now they are conducted by someone else.

    • another example: last year I ran two one-day nonviolent communication trainings, one guy got interested, volunteered to run NVC trainings weekly, and have been doing so for more than a year since then

  • I’ve put a lot of effort into our internal events management system; it automates the postings to 3 online services (Facebook, VK, Timepad), generates announcement images, collects the attendance statistics through the Slack bot, etc. It allowed us to cut down the time spent on each event from 20–30 minutes per event to ~5 minutes.

  • We usually have a dedicated person who invites external speakers (one or two talks per week); these talks let us attract a non-LW-but-adjacent crowd (first step of the funnel), some of which get interested and join the rationalist community later.

Rationality workshops

Running the workshops was our original goal for Kocherga.

We assembled our curriculum from various sources, but mostly from the posts on CFAR techniques and working through the original research by ourselves.

I finally got my hands on CFAR handbook from one of the CFAR alumni last winter and was satisfied with how much we got “right” (i.e., there are no contradictions between what we teach and what CFAR teaches, and the material covers the same topics). Our workshops are typically two-days long, so they’re more condensed and don’t cover everything. OTOH, we do include Bayes theorem (I heard that CFAR doesn’t) and I believe we cover VNM and prospect theory in more detail, but I can’t be sure about that since I didn’t attend CFAR workshops myself.

Our current set of classes is:

  • Day 1:

    • Goal Factoring

    • Inner Simulator

    • Premortems

    • Expected Utility (VNM, prospect theory)

    • bonus: Resolve Cycles

  • Day 2:

    • TAPs and Deliberate Practice

    • Aversion Factoring

    • bonus: GTD

    • Internal Double Crux

    • Probabilities (reference groups, Bayes theorem)

    • bonus: Recursion (i.e. “let’s apply rationality skills to our task of aquiring rationality skills”)

The attendance at workshops varies from 6 to 15 people. We have a scholarship program and take 1–2 students per workshop for free or with a significant discount. We also allow our staff members to visit workshops for free when the situation allows it.

Some data:

  • 2016: 4 workshops in Moscow.

  • 2017:

    • 6 workshops

    • 1 more workshop in Saint-Petersburg

    • 1 three-week course (one day per week + homework assignments)

    • 3 corporate workshops

    • a summer camp school with an intensive two-week program

  • 2018 so far:

    • 2 three-week courses

    • 4 workshops

    • one experimental “rationality for executives” course

    • 1 corporate workshop

The basic price of a workshop is $320 (20,000 rubles) for two days. We’ve doubled it this year (from $160) and didn’t notice a significant drop in sales, but we probably can’t go much higher.

Visitor trackings and other automations

We offer visitors our club membership cards. These are not mandatory (you take a guest card if you don’t have a club member card), but holding a club card gives you a small discount (via yudcoins which we give distribute among club members on every visit), while allowing us to track each visitor by their personal id. I’m planning to analyze the retention statistics based on this data, but haven’t got around to it yet.

For the per-event statistics I wrote the Slack bot which asks a receptionist for the number of people attending an event and stores it in our database.

I also wrote a few custom online pages for the visitors:

Other infrastructure

Besides the already mentioned online services, event management system and visitors tracking, we had to solve many other infrastructure tasks:

We have an internal wiki (mediawiki) with the documentation on everything, from “how to fix the coffee machine” to “checklist for announcing new workshops”.

We’ve got a basic hang on most of our legal and financial operations — taxes, recurring payments, labor contracts, accepting online payments.

We’re managing a mailing list and two social network accounts (Facebook and VK.com — both are popular in Russia).

We’re also probably the only anticafe in the world with a public API.

Failures and key issues

Physical space

We’re still understaffed and under-financed for maintaining our physical space. There are many small issues (ragged couches, wall paint issues, old lightbulbs) which we’re constantly trying to stay on top of, but sometimes we lack cash and sometimes we lack person-hours to fix it all.

Our largest issue with the location is air conditioning, which is barely tolerable, and we’re probably losing some people because of it (what’s worse, we’re in a negative feedback loop situation where whenever we get more visitors than usual they notice the air conditioning issue more often). The project to upgrade the air conditioning would cost around $5000 and we can’t afford it right now, unfortunately.

The second largest issue is that our biggest room can only fit 40 people. We have to reject some potentially useful events (larger conferences, talks) because of this.

Our lease is year to year and the landlord is selling the location (but they’re not in a hurry and the new owner would hopefully be ok with continuing to rent it to us). This is an existential risk which we can’t satisfactorily solve right now — we can’t afford to move and we can’t buy this location ourselves (we’d love to, actually, but the price is too high for a mortgage).

Profitability

Anticafe is a low profitability business even in the best case scenario. A usual cafe can serve a client in 1–2 hours and get $10–$20 in return, so their rate is $10/​hour/​client, while anticafe’s hourly fee is $2–$2.5/​hour. Cafes and anticafes are competing for the same city space, so the rent is naturally reaches the equilibria where it is just low enough for the businesses which can stay profitable to continue existing.

As far as I know, some other anticafes solve their profitability issues by entering the race to the bottom — drop the price as low as you can, stuff as many high school students as possible in the same room, get a hold on their attention with video and board games, it’ll be loud and overcrowded, but you’ll survive. While you’re engaging in this process, throw under the bus all your expectations about your target audience and about what kind of community you’re trying to build, because if people are asking for FIFA on PlayStation, give them FIFA and shut up about your boring rationality topics.

Ok, I’m not saying that this Moloch menace is the main reason why we’re not yet consistently profitable: we’re not yet overcrowded, we could do a better work of building a community faster, we could work more and be rewarded with more attention from those who are a good culture fit for Kocherga (we definitely don’t have market saturation issues yet, Moscow is a big city). I’m just saying that we had to be careful with the tradeoff between quicker growth and keeping our values intact.

I believe we could reach the positive cash flow in a few months, but I also know I was overconfident about this last year and the year before that, so I distrust my own predictions on this.

Financial breakdown

Expenses per month:

  • Rent: $3500

  • Cleaning services: $300

  • Salaries (several receptionists and one full-time manager): $1700

  • Supplies (cookies, coffee, tea, milk, paper, whiteboard markers, etc.): $800

  • Taxes: $900 on average (varies from month to month, I’m not sure how precise this number is)

  • My personal expenses: $1000, half of which is rent (I don’t pay salary to myself because I’m a sole proprietor; this also means that I’m legally accountable for all our finances with everything I have and own, which fortunately isn’t much anyway)

  • Ongoing repairs and occasionally buying new stuff: varies from month to month; let’s say $500

Total expenses: around $8700 on average; the actual amount is something like $7500–$10000 with 80% confidence.

Revenue:

  • Pay-per-time revenue: $5000–$5500 (lower in summer, higher in winter; July 2018 was our best month ever, which is a good sign)

  • Rationality workshops and other trainings: $1500–$3500 depending on our trainings schedule.

Total revenue: $6000–$9000.

So, as you can see, we’ve been losing from $1000 to $3000 per month. This gap has been decreasing over time (a year ago I considered a -$3000 month to be bad and -$1500 month to be normal; right now -$1500 is bad and from -$500 is ok), but we’re almost out of money. Actually, depending on the time of the month, sometimes we’re completely out of money.

September is nigh. September can bring us a large influx of visitors. Especially considering that Russian translation of HPMoR will go into print and we’re the main pick-up point for it. There’s a crowdfunding campaign going on for printing HPMoR’s Russian translation which raised $70000 so far; HPMoR is quite popular in Russia.

But I’m not sure we can make it without bankrupting — we’re very financially vulnerable right now and any unexpected emergency can blow us off our path. And, of course, working with zero cash runway is frustrating and wasteful — we can’t invest into any long-term projects, we can’t afford salaries for talented community members which would be up for the job but have to work on something else instead, we can’t outsource everyday errands to free ourselves for more important work.

Our future plans (assuming we stay afloat)

Managing a business is a full time job and often it’s a distraction from our main mission of raising the sanity waterline and growing the community. Someone had to do it, and I still believe this is was a good investment of time and effort. I’d be lying if I said I didn’t enjoy it — building an infrastructure for the community to enjoy is very satisfying. We’ve changed the lives of many people for the better; people meet their partners, change their careers, find jobs, read Sequences, discuss and update their beliefs because of us.

Still, building a happy community is not the goal. Changing the world is. We should expand much further, improve our rationality curriculum, work on teaching skills, write blog posts and books (Pion is writing a book currently; I’d love to do the same but I’m constantly overwhelmed with all this business management work), do some serious research, stir up more activity in the larger rationalist community in Russia, and so on.

More specifically, in the next few years I hope we’ll:

  • Expand our instructors team to 8–10 people; train 300–500 students per year.

  • Open a few more rationality spaces in other cities in Russia. (This would be tough, but our infrastructure is solid and it makes sense to try to scale it up.)

  • Find a new venue in Moscow with more space, air conditioning and natural lighting. (Remember, we’re currently located in a basement.)

  • Increase our brand recognition tenfold or more.

  • Publish a few books, release an online rationality course, make our workshops more effective and impressive.

  • Engage the local rationalist community more, sponsor the talented members with targeted grants which would allow them to produce more content and help the movement.

  • Give back to the English-speaking rationalist community — post more on LessWrong.com, share our experiences and successes.

  • Figure out how to approach rationality research — both on how affect people’s lives (a la CFAR longitudal study) and on the core rationality topics and on how to improve the state-of-the-art approaches.

How much do we need?

Here are a few reasons why donating to Kocherga might be a good idea:

  • We’re currently funding-constrained, not talent-constrained.

  • Rent, salaries and taxes in Moscow are significantly lower, so impact-per-dollar here would be higher than in Bay Area or New York or most other large US cities.

  • Counterfactually, losing Kocherga would be a big loss of utility for a lot of people.

  • Our existence is probably making a positive impact even for the rationalist community in the US, simply by increasing the number of people who get interested in rationality and become involved in global EA activities, AI safety or even relocate to USA (it’s hard to measure, though).

How much do we need?

  • $1000/​month would significantly increase our chances of staying afloat.

  • $2000/​month would mean that we can be confident about staying afloat, as well as allow us to improve and grow faster (by hiring one more full time staff member or by hiring one part-time staff member and directing extra cash to renovations); the only risk would be existential about being evicted by the landlord or about the venue being sold to someone who doesn’t want to rent it to us.

  • at $3000/​month or higher we’d start looking for a new venue in a few months or consider a mortgage for the current venue.

Note that all these numbers are crude estimations and don’t take into account Patreon fees and Russian taxes (6% flat rate).

Extra option: one-time sum in a range of $100k–250k would allow us to buy the current venue, which would have a huge impact on everything else. ($250k would cover the full price of the venue; $100k is the lower bound after which I’d seriously consider a mortgage. Owning the venue would also mean that we could invest in a major overhaul without worrying about long-term returns.)

Ways to donate

We’re launching a Patreon page today.

Alternatives:

berekuk25 Aug 2018 0:28 UTC
171 points
23 commentsLW link