Epistemic status: Not a historian of science, but I have thought fairly extensively about meetups. Kind of making this up as I go along, almost certainly missing important points.

Other meta: Written all in one sitting, to not let the perfect be the enemy of the good. No one proofread it, so hopefully there aren't sentences that just cut off in the middle. Also, forgive my excessive use of scare quotes.

tl;dr: The difference between historical salons and LW meetups is that meetups do not feel like the place where progress is made. They’re not doing research or publishing anything. If they wrote up the results of their discussions and posted them on LW, that would help conversations be less siloed.

This post is a confluence of two major topics I’ve been mulling over for a while now.

Number one is rationality meetups. I’ve been working on global meetup coordination on-and-off and more-or-less formally for over two years now, and it is just... not an easy problem. It was about two years ago that I wrote What Are Meetups Actually Trying to Accomplish?, and I honestly haven’t made a ton of progress since then (due to meetups being like my sixth or lower priority at any given time). However, I have changed my outlook somewhat - that post was very much trying to be descriptive, but I’m now much more inclined towards prescriptivism (more on that in the appendix).

Number two, Habryka has gotten me hooked on books about the history of science, and on thinking about how intellectual progress is made more generally. I’ve recently read The Innovators, The Age of Innovation, and the biographies of Steve Jobs and Elon Musk, and I’m in the process of reading The Making of the Atomic Bomb, Science in the 20th Century, and a biography of Benjamin Franklin. I also majored in physics, which de facto included a fair amount of history of science.

Turns out, most modern histories of science and innovation (at least the ones that I’ve read, but like, selection bias) focus pretty heavily on the first half of the 20th century, and specifically, the people during that time who roughly thought of themselves as physicists.

Physics at the time was a small community. If the books I’ve been reading are to be believed, all the big-name scientists read each others' papers, many of them were close friends who regularly exchanged letters, and they all went to lectures by each other and often had their big ideas there.

I’ve come to think of the thing that meetups should be as something akin to these lectures (but mixed with intellectual salons / Franklin’s Junto / gentlemen’s clubs). A version of meetups I’d be excited about would be not only forums for exchange of ideas, but also ones that aim towards real intellectual progress.

Glossing over all the detail and individual variation, meetups are currently a mix of a) social groups, and b) forums for learning, intellectual discussion, and exchange of ideas. For the latter, we generally have a core set of ideas that we’re all more or less familiar with, and we build from there. This is much like an intellectual salon or Franklin's Junto. However, I think there are a few key things missing that prevent us from developing and building on ideas in a way similar to early 20th century scientists:

Missing pieces

Less clear focus

I have a sense that, compared to early 20th century scientists, our efforts are less focused on one thing in particular. ‘Rationality’ is a broad umbrella, encompassing AI safety, developing the art of human rationality, navel-gazey community stuff (like this post), science more generally, and everything that falls under the ‘effective altruism’ umbrella (animal welfare, global development, cause prioritization, etc, etc, etc.). This means that even people who are really trying to ‘make progress’ on ‘the project of rationality’ are often working on completely different things, and won’t even be able to talk to each other in depth about what they’re doing.

However, this isn’t necessarily a dealbreaker - it’s similar to the way that particle physicists and astrophysicists will share a lot of basic knowledge, but there’s still plenty of detail about their research that they may not easily be able to share with each other.

Feeling of license

I’ve gotten a feeling - both from talking to meetup attendees who don’t live in ‘rationalist hubs’, and from being one myself - that most meetup attendees have a feeling that they’re ‘not part of the intellectual elite.’ Speaking for myself, there was a feeling that all the ‘real’ progress was happening in conversations in the Bay Area that are never written down, and that there were these special elite people like Eliezer, Scott, Nate Soares, Luke Muehlhauser, Julia Galef, Anna Salamon, etc., who were the only ‘real rationalists’, and other people just didn't really get to contribute to rationality. This resulted in a feeling that because I wasn't part of this select elite, I wasn't smart enough to be working on rationality, and so it wasn't worth writing down and developing my ideas.

Sense that it matters

As gruesome and horrible as the World Wars were in terms of immediate human cost, history of science books make it pretty clear that they were really good for science and progress. There was a sense of urgency, coupled with plenty of government funding, and this resulted in innovations like the transistor and the radar, as well as huge developments in fields like nuclear physics.

Many people in this community do have a sense of urgency, because of existential risk. But I think it’s fair to say that AI safety, for most people, feels less viscerally real than the current, imminent, or past gruesome deaths of their friends and family members. In WWII, people knew what the end of their world looked like, because it was happening all around them. For us, we may have a sense that our world is very close to ending, but it usually feels pretty intellectual. (I think events like Petrov Day and the Secular Solstice are valuable for bringing existential risk into near mode, but that's another story).

Ending death is definitely more visceral, but it’s still very different from the World Wars. During that time (at least as the history books tell it, so could be way off), most of the Western world was pretty focused on a single goal. By contrast, ending death is currently a pretty out-there idea, and it's much harder to rally behind a cause when people think you're crazy than it is when multiple countries are on your side.

Siloed conversations

Conversations in the geographically disparate communities are fairly siloed. This means that, while past-Mingyuan all the way over in Chicago may feel left out because she doesn’t know what’s being discussed in Berkeley, the flip side is that the people in Berkeley have no idea what’s being discussed in Chicago.

As a result, different meetups are probably endlessly retreading much of the same ground. While this isn’t always a bad thing - since discussions like those can be useful just for practicing things like debate and idea generation - it does feel like there’s a lot of wasted motion. There must have been dozens of meetups on Meditations on Moloch, and I’m sure most of them have had plenty of ideas and insights, but there’s no net progress because we don’t communicate with each other.

How can we be more like early 20th century science?

I have several concrete ideas for improving on the current situation, most of which I’ve sanity-checked with Habryka. Some of these are going into effect immediately, while others are more long-term goals that we aren’t currently prioritizing.

Fostering a feeling of license

Habryka and I have talked a bit about how to make meetups feel more like real institutions, rather than just random groups of people. One idea we’ve floated is having there be ‘official’ LessWrong groups - similar to how other organizations (Toastmasters, Girl Scouts, etc.) have official chapters. While I think this would probably be cool, it would also require a lot more work than any of us currently have the time to put in.

Something that we can do in the short term is build a feature to allow people to make frontpage LessWrong posts that are affiliated with meetup groups. I’m not sure how much this will actually foster a feeling of license, but at least it will make my other proposal work better.

A single conversational locus

Three years ago, Anna Salamon wrote On the importance of Less Wrong, or another single conversational locus. In direct correspondence to her point about the online rationalist diaspora, I think the various geographic communities could benefit strongly from having a more centralized discussion.

Proposal: Monetary incentives for write-ups

I propose that meetup attendees write up the ideas that come out of each meeting - or even just summarize the conversation they had - and post them on LessWrong. In order to encourage people to actually write these posts, I am offering a monetary incentive of $100 per write-up (to be paid out a maximum of once per week to each meetup). See this doc for further details.

Proposal: Facebook group

I’m also in the process of creating a Facebook group for attendees of all meetups worldwide. The goal is to make meetup attendees feel more like part of a single community, and to have a Schelling place for them to discuss problems specific to meetups.

I've wanted something like this to exist for a while, and in a survey I ran a couple months ago, more than 80% of respondents indicated that they'd appreciate a way to communicate with people from other meetup groups. Of those, a plurality preferred to use a Facebook group, and much as I personally dislike Facebook, I have to admit that it has pretty good infrastructure for this use case.

Some other ideas:

Cross-pollination of ideas via geographical travel.

Obviously this won’t be feasible for everyone, or even most people. However, I’ve found that there’s value to meeting people face-to-face that’s really hard to get even with years of online discussion. This is why a bunch of people were excited to accompany Scott Alexander on his Meetups Everywhere road trip - it’s just really hard to know what’s going on in a place without actually going there.

Topics of the month.

I’d potentially like to generate a default meetup topic for each month, so that lots of people all over are talking about the same thing at the same time. This could make discussion on both LessWrong (via the write-ups) and the Facebook group more focused, and thus hopefully move the community's conversation around those topics forward (trying to avoid the failure mode of retreading the same ground).
I am not going ahead with implementing this idea imminently both because I'm less confident in it, and because of the overhead required.

Conclusion

So, I have some ideas and some amount of power to execute them. I am fairly excited about the write-up incentives idea and fairly confident that the Facebook group is at least something that enough people want to try. On the other hand, many of the times that I've previously tried to publicly take action in this space, I got shot down pretty harshly. So I'd love to hear why you think my proposals are naïve / misguided / going to pollute the commons / going to crash and burn before I launch anything this time. Concrete critiques and proposals of alternatives would be greatly appreciated.

Appendix: Towards prescriptivism

So, why the change from the descriptive approach in my last post to a more prescriptive outlook? Well, frankly, it became pretty clear to me that there had been barely any thought put into meetups as a whole (not to diminish the work that has been done in this domain; there’s just been very little of it overall). Meetups kind of grew organically because people wanted to socialize with people who shared their interests, but over the last decade or so, many of them have grown into institutions that do tangible things like invent holidays, run their own rationality workshops, and translate the Sequences into different languages. While all of those things are really cool and great, the majority of meetup groups aren't doing anything like that. So I've come to think of meetups, as a whole, as vastly underutilized resources. If this is going to change, someone needs to take real action on it, and I've somehow ended up in a position where I guess that's my job. So here we are.