A quick strategy question for you. You mentioned that
1) Most of the impact comes from the largest groups
2) Local Groups risk death or dormancy due to loneliness
Given 1), I’m interested in why you’ve decided to work on interventions for 2). It seems plausible that a better use of resources would be to double down on the large groups and seed new groups mostly for the purposes of seeing if they get large.
Thanks for the question, although I’m sad that you didn’t include my jokes… ;)
I had this in an earlier iteration of the post (I’d cut it for the sake of brevity) so I’ll just paste here with some additions:
“There will be comparatively more time spent on converting medium-sized groups into large groups: we know that most of the impact so far has come from this. [In addition, focusing on city group growth is more likely to be counterfactual than focusing on university group growth]
So why spend any time at all on making sure that groups don’t die out? Well, we want to grow the community, but there is also an abundance of value in:
a) Making the community welcoming
b) Building a diverse community
c) Strengthening the commitment of existing EAs
...and thus having an overall greater impact.
If there is a recurring issue that people feel lonely or unwelcomed by the community [see the EA Survey and the Local Group survey], this is something we need to address for the long-term impact of EA.
Likewise, if group growth is solely dependent on pre-existing hospitable conditions and no one is helping the less likely candidates, not only do we reduce our chances of having a counterfactual impact, we’ll also miss out on a wealth of viewpoints that represent valuable opportunities for impact. [As well as making local group work more susceptible to the Schelling Effect.]”
On seeding new groups and seeing if they get large vs improving our services for the hundreds of groups we’ve already seeded:
Seeding groups and simply “seeing if they get large” would be ok—it would mean we nudged them to take the first step—but it wouldn’t be great. The point is that, based on what we’ve observed from the hundreds of groups we’ve seeded, we think they wouldn’t “get large” without our intervention. We think most of them would die out, so our work to provide encouragement and guidance is extremely valuable.
Thanks for this.However, I’m still a bit confused.
You seem to be saying that devoting resources to smaller groups is useful for:
a) Making the community welcoming b) Building a diverse community c) Strengthening the commitment of existing EAs
But, presumably, all of these goals are better accomplished by spending resources on the groups with the most people since all of these benefits depend on the number of people in the groups.
It could be that you think preventing groups from dying out is important for improving diversity in the community, but this could only be true if some demographics are systematically more likely to trying to start an EA group and fail. Otherwise, we would expect the demographics of chapter leaders to roughly match the demographics of EA as a whole.
The reason I’m asking all of this is to try to compare models of why groups succeed and why groups fail. Presumably, our goal is to create vibrant, self-sustaining communities like those in SF, London, Boston and elsewhere. I have a lot less on-the-ground experience than you do, but I think there are roughly four phases that a group needs to go through to be effective and sustainable.
1) The group is founded by (or quickly attracts) a highly dedicated, energetic, and skilled founder.
2) The founder is able to attract a small group of highly dedicated chapter members to help found the group.
3) The group establishes tactics for regularly attracting new chapter members.
4) The founder sets up a system for passing leadership of the group onto a highly dedicated chapter member.
I think the skill of the founder matters a great deal because getting people to join a chapter is much harder early on than it is later. This is because chapters are more valuable the more people are in them. We should expect that recruiting the second person is going to be much harder than recruiting the 20th person.
My rough hypothesis is that we probably lack the ability to substantially improve the skills and abilities of founders and so the supply of excellent founders is probably the bottleneck on establishing sustainable groups.
If this is true, then it makes sense to spend time identifying and cultivating highly-dedicated founders, but it probably doesn’t make sense to spend much time trying to save groups from disappearing. I’m interested in whether your agree with this model and if not, where you think we disagree.
Thanks for the post.
A quick strategy question for you. You mentioned that
Given 1), I’m interested in why you’ve decided to work on interventions for 2). It seems plausible that a better use of resources would be to double down on the large groups and seed new groups mostly for the purposes of seeing if they get large.
Thanks for the question, although I’m sad that you didn’t include my jokes… ;)
I had this in an earlier iteration of the post (I’d cut it for the sake of brevity) so I’ll just paste here with some additions:
“There will be comparatively more time spent on converting medium-sized groups into large groups: we know that most of the impact so far has come from this. [In addition, focusing on city group growth is more likely to be counterfactual than focusing on university group growth]
So why spend any time at all on making sure that groups don’t die out? Well, we want to grow the community, but there is also an abundance of value in:
a) Making the community welcoming b) Building a diverse community c) Strengthening the commitment of existing EAs
...and thus having an overall greater impact.
If there is a recurring issue that people feel lonely or unwelcomed by the community [see the EA Survey and the Local Group survey], this is something we need to address for the long-term impact of EA.
Likewise, if group growth is solely dependent on pre-existing hospitable conditions and no one is helping the less likely candidates, not only do we reduce our chances of having a counterfactual impact, we’ll also miss out on a wealth of viewpoints that represent valuable opportunities for impact. [As well as making local group work more susceptible to the Schelling Effect.]”
On seeding new groups and seeing if they get large vs improving our services for the hundreds of groups we’ve already seeded:
Seeding groups and simply “seeing if they get large” would be ok—it would mean we nudged them to take the first step—but it wouldn’t be great. The point is that, based on what we’ve observed from the hundreds of groups we’ve seeded, we think they wouldn’t “get large” without our intervention. We think most of them would die out, so our work to provide encouragement and guidance is extremely valuable.
Thanks for this.However, I’m still a bit confused.
You seem to be saying that devoting resources to smaller groups is useful for:
But, presumably, all of these goals are better accomplished by spending resources on the groups with the most people since all of these benefits depend on the number of people in the groups.
It could be that you think preventing groups from dying out is important for improving diversity in the community, but this could only be true if some demographics are systematically more likely to trying to start an EA group and fail. Otherwise, we would expect the demographics of chapter leaders to roughly match the demographics of EA as a whole.
The reason I’m asking all of this is to try to compare models of why groups succeed and why groups fail. Presumably, our goal is to create vibrant, self-sustaining communities like those in SF, London, Boston and elsewhere. I have a lot less on-the-ground experience than you do, but I think there are roughly four phases that a group needs to go through to be effective and sustainable.
1) The group is founded by (or quickly attracts) a highly dedicated, energetic, and skilled founder. 2) The founder is able to attract a small group of highly dedicated chapter members to help found the group. 3) The group establishes tactics for regularly attracting new chapter members. 4) The founder sets up a system for passing leadership of the group onto a highly dedicated chapter member.
I think the skill of the founder matters a great deal because getting people to join a chapter is much harder early on than it is later. This is because chapters are more valuable the more people are in them. We should expect that recruiting the second person is going to be much harder than recruiting the 20th person.
My rough hypothesis is that we probably lack the ability to substantially improve the skills and abilities of founders and so the supply of excellent founders is probably the bottleneck on establishing sustainable groups.
If this is true, then it makes sense to spend time identifying and cultivating highly-dedicated founders, but it probably doesn’t make sense to spend much time trying to save groups from disappearing. I’m interested in whether your agree with this model and if not, where you think we disagree.
Thanks for this, I don’t think we disagree all that much actually—let’s chat about it in our Skype.
And for anyone else that wants to chat: calendly.com/georgiedotimpact