While other groups might find their local reality to differ substantially, we (the EA Geneva team) hope to convey some useful general insights by sharing what our community building efforts look like nowadays, and why.
Our experience comes from three years of building up:
- A small local community next to our studies (2015-2017); and
- An increasingly diverse association with full-time staff since July 2017.
This first post is structured as follows:
- The engagement funnel
- Activity details
2. The engagement funnel
This post documents our current model of building a local EA community.
Our funnel captures the essential activities that have shown promise in fostering people’s journey into EA. We have not reached our current ideal design, though, and will likely make changes to it in the future.
Generally, the funnel serves two goals:
- Impact: realise our local community’s potential on a global level (mainly by encouraging people to go into top priority roles); and
- Health: build a locally anchored community that is exemplary of the guiding principles of EA (mainly through personal interaction, self-development & collective thought development).
2.1 - Key models for community tactics
As highlighted through the craft is not the community, actually causing externally verifiable impact often feels like it interferes with serving the internal needs of the group, especially when starting out. Developing explicit models of our key audiences and their paths through the funnel has helped us to work out a somewhat balanced two-pronged, goal-driven approach.
Our key audiences are:
- Students (undergrads & graduates)
- Early career professionals (PhDs, postdocs, trainees, junior positions)
- Professionals (35+ y.o.)
We try to identify valuable modes of engagement to contribute to people’s development depending on their
- Position in the three-factor model;
- Personal situation; and
- Length of stay in Geneva.
We see three general development stages that our activities have to serve:
- Attraction: attracting people with high impact potential;
- Education: developing their knowledge and skills; and
- Engagement: consolidating their paths to contribution.
They theoretically appear to be sequenced but we found they are tightly interrelated in practice. Most of our activities nowadays serve more than one stage, as you will see below.
In parallel, two meta activities are part of our community building:
- Local coordination: coordinating local activities; and
- Global integration: tying everything into the global community.
Local coordination is the subject of post 4 in this series. Global integration deserves an entirely separate discussion.
2.2 - Segmenting the community
To keep segments relevant yet simple, we decided on three concentric circles:
- Public: anyone (potentially) interested;
- Member: individuals having completed our introduction workshop; and
- Core: individuals having completed our advanced workshop.
One can join the member segment without a workshop if two core members, after a 1-on-1 each, judge that one has demonstrated full knowledge of the subjects of the intro curriculum. We do not currently plan to make such exceptions for the advanced workshop.
To realise our community’s potential, the funnel is thought to actively drive people towards
- A better understanding of the world and EA;
- Growing engagement and responsibilities within the community; and
- Increasing access to resources, activities and infrastructure.
We expect this way of escalating asks and rewards to result in more meaningful activities for all segments and to motivate people to move further down the funnel. The key drivers from segment to segment are the workshops.
- Publicly accessible activities:
- Themed socials
- Introduction workshop
- Introduction talks to promising audiences
- Ad hoc events (often in collaboration with other organisations)
- Members and core:
- Working groups
- Members-only themed socials
- Access to internal communications platform
- Access to educational resources
- Possibility for peer mentoring
- Access to community coworking and hangout space
- Self-improvement group
- Workshop facilitator training
- Discussion dinners
2.3 - The minimum viable funnel
Thinking back to the beginnings of EA Geneva, we expect that we would have learned a lot faster if we had exclusively focused on building up the core activities that we, today, perceive as the backbone of the funnel.
Especially in the beginning, meaningfully broadening the range of our activities proved difficult and most of it seems to have been a waste of time in hindsight. Developing the capacity to maintain this core funnel structure would have:
- forced us to focus on learning about all the key topics; and
- given us a set of activities allowing to integrate new contributors more effectively.
From the start, 1-on-1s seem among the most effective ways to identify potential core members, address doubts, establish common knowledge, and develop someone’s mode of contribution.
To build the knowledge, experience and skills necessary to run public workshops, we think it can be useful to start off by using the themed socials to study the subjects of the curricula of the intro workshop and advanced workshop (linked in the descriptions below). Depending on what the core group feels comfortable with, these exploratory sessions could be public, invite-only or somewhere in-between. For very small groups, the curricula could also be content of 1-on-1s.
Upon sign up to an intro workshop, sending out EA reading recommendations helps us to further automate selection with two main purposes:
- Forestalling that people come to the events with faulty ideas about EA (e.g. interpreting EA to mean “helping where you can see the impact with your own eyes”); and
- Making sure people are sufficiently proactive and open-minded such that a session is productive (people commit to 30 min of reading when signing up).
Presenting the group, its goals, its activities, and its principles explicitly at the beginning of each public event also helps people to identify with the group - or not; motivates new contributors; and solidifies the group identity.
To complement these key efforts, explicitly making the time for regular self-improvement group meetings to work on the core members’ personal development helps to set the tone for the broader community culture. Proactively sharing and discussing feedback on room for personal improvements, group dynamics and organisational optimisation has proven invaluable to us as individuals, as a team, and for the community. For us, these sessions will stay invite-only to preserve the necessary intimacy needed to work on personal bottlenecks together.
3. Activity details
These are all the activities that we are currently organising as parts of our funnel:
Below, we detail them out. Priorities might change according to various considerations, among them:
- Evolving activities: some formats (can/should) change with the maturity of a group (e.g. introduction workshops: from self-study to public-facing workshop); and
- Group-dependent activities: adaptations based on e.g. specialities, needs, interests, and local strategy.
- Advance someone’s career planning;
- Advance someone’s understanding of EA-related topics;
- Advance someone’s involvement in the local community; and/or
- Work on applied rationality skills or other self-development.
- In-person meeting between two people with:
- A clearly defined objective in line with the possible purposes.
- Predetermined duration of 30min to 1h, depending on the utility;
- In a relaxed location with:
- Sufficient privacy to talk about personal things;
- Tables to take notes on; and
- Focus on individual interests, bottlenecks and potentials;
- Take notes of meetings and log them in a tracking system to analyse it later;
- If not too costly, mix 1-on-1 partners up every now and then;
- Make sure core group members also spend 1-on-1 time with one another.
- Create a Schelling point for newcomers and members by:
- Offering valuable content for members; whose presence allows
- Mingling with newcomers to present the community welcomingly.
- Making space for members to contribute by sharing their knowledge.
- Quiet location with drinks;
- Intro & EA pitch (10min);
- Themed contribution (1-2h):
- Structured activity based on a contribution from community member (sharing their work / a read / an experience / a question / a skill);
- Wide range of topics and formats possible (moderated discussion / practicing or studying / talk with Q&A / etc.).
- Socialising time (1h+);
- Social time to chat, exchange and foster the community ties.
- If the topic is easily accessible and directly related to EA, socials are public;
- If the topic is “too far out” or technical, we aim it at a specific audience and adapt the location;
- Experimenting with formats, topics, location and audience allows for more targeted outreach one event at a time;
- Topics should be in the interest of at least a subset of the community to have a decent amount of members present;
- Once established, this is a very low-cost activity - we run this format biweekly.
- Get interested individuals to become actively engaged with the matter;
- Offer a welcoming and representative public activity without over-simplifications;
- Get to know interested people to explore their potential.
- A 6 hour curriculum with exercises and discussions, including:
- Raisons d’être of EA and EA principles;
- Cause-prioritisation and cost-benefit analyses;
- Consequentialism, values, and beneficiaries;
- The human mind and human(e) rationality;
- The global EA community, its successes & current opportunities;
- Career planning, donating, and promoting the ideas.
- Curriculum has 6 modules that could be given by different facilitators;
- We have run two formats:
- 3 week seminar with 2h-sessions once per week;
- 8h-workshop with collective lunch (we have developed a slight preference for this due to the group dynamics resulting from its intensity).
- Other remarks
- We have had between 10-16 participants each time and recommend 28 as upper bound;
- A good facilitator to participant ratio seems to be 1:6;
- We’d like to see experimentation with other formats and exercises;
- We try to ensure high fidelity idea spreading by:
- Leaving room for deviations from prepared content; and
- Not posing as an expert group but as individuals who are still learning about it ourselves.
- The setting is ideal to get to know new additions to the community - make sure you register people’s interests, needs and potential somewhere to motivate and integrate them optimally.
- Establish a solid understanding of advanced EA-related content and equip members with philosophical and methodological knowledge;
- Get members to a level where they can contribute to the local and global community’s intellectual development and quality.
- Curriculum covering:
- Idealised ethical decision-making;
- Practical ethical decision-making;
- Coordination and community.
- We will run this exclusively as a 1-day format;
- Very discussion- and exploration-heavy - fewer presenting than in the intro WS, more guided by questions.
- Other remarks
- We have not yet run this workshop and expect many changes to come;
- To ensure productive discussions, we will only invite people who have had positive 1-on-1s with at least two core members;
- See introduction workshop above.
- Secure long term involvement by translating (dormant) potential into (impactful) action;
- Build intellectual vibrancy by fostering opportunities for advanced intellectual engagement;
- Create reports and project proposals based on in-depth explorations of a chosen topics to foster understanding and goal-orientation of the local community.
- Once 2 or more core members are willing to take the lead, a working group can get started;
- Each group needs a detailed project plan with goals and metrics;
- Members can suggest and join working groups but not lead them;
- Groups are mainly self-organised but group leaders send weekly updates and attend monthly governance meetings.
- Other remarks
- Choice of leadership, cause-area and approach are of great importance;
- Has to fit into the broader local community building strategy;
- Make sure to have the capacity to give adequate support;
- Potential downside risks need to be analysed and avoided;
- For more, read here.
Community socials and discussion dinners
- Promote community spirit through bonding activities
- The following formats can be combined with one another:
- Members-only themed socials;
- Discussion dinners;
- Parties; or
- Other remarks
- Depending on the activity, sometimes only a limited amount of people can attend - some activities should even be invite-only - like discussion dinners.
- Make sure each activity has meaning to it;
- As it is members-only, less EA-related topics are welcome, too;
- Organise content-free activities only around once a quarter;- a plain BBQ is only interesting if everybody knows each other sufficiently well.
Targeted talks and co-organised events
- Gain visibility among new, perceptive audiences
- Tailored to a specific audience
- EA speaker or panelist
- Talk with Q&A or panel discussion
- Offer aperitif for mingling afterwards and to talk to promising interessees
- Can be a big public event (e.g. inviting Peter Singer) or simply an introduction talk for a partner organisation (e.g. the local Rotary, LIONS Club, or Mensa chapter);
- Do not expect members to switch sides after your talk - the goal here should be to win allies for cross promotion.
- For intro talks, have one or two people who can always give the same 20 min presentation and then professionally handle a long Q&A or moderated discussion;
- If collaborators are chosen wisely, these events can be good to gain visibility with relatively small investment;
- Events solely for visibility seem pointless beyond practice or entertainment reasons - make sure the audience is receptive and the setting allows for personal chats afterwards;
- Do not expect to initiate collaborations - this has turned out to be hard. A good outcome is 1-2 new active members per event.
We see sub-communities as a subset of the fractal that is our local group, which is a subset of the fractal that is the global community. We are still unsure about the best conditions to start these but think that any community above some critical mass of active core members will likely benefit from regrouping.
- Make space for members with different profiles to build their own mode of engagement and contribution
- Foster the right kinds of diversity under a common goal
- Have to be lead by at least two core members;
- Have to integrate into the organisational structure;
- Can organise any of the activities outlined above for their audience;
- Can be organised around a topic or specific demographic;
- If topical, events can be organised by the group for any audience, depending on the content and format;
- If demographic, the group will likely benefit from staying exclusive.
- Avoid fracturing the community: sub-communities should not tear apart the existing community but make space for more engagement by clearly defined subsets of our members;
- Choice of coordination mechanisms, leadership, focus and approach are of great importance;
- Has to fit into the broader local community building strategy.
- Groups will become a force of themselves within the local community;
- Ensure a maximum of coordination from the start;
- Make sure to have the capacity to give adequate support and foster integration;
- Potential downside risks need to be made explicit.
- Student communities
- Uni of Geneva
- Communities of specific demographics
- 35+yo professionals
- Residents of Lausanne
- Communities of common interests
- AI Safety
Self-improvement for world-improvement community
A slightly special sub-community that we think deserves more attention, is a self-improvement community. This means, expanding the exclusive self-improvement group(s) to the entire community. From where we are standing, it looks like the only way the EA community will have long-lasting impact, or will itself last the test of time, is if people find a tribe within it. Self-care, cultivating compassion and applied rationality, we hope, might help to build a basis that does not just create further in-group vs. out-group dynamics.
- Establish a self-improvement community
- Examples of regular activities:
- Structured feedback
- AR tutoring wheel
- AR technique revision
- Individual debugging
- Pair debugging
- Examples of on-going activities:
- Prediction markets
- Accountability pairing
Co-written by Nora Ammann and Konrad Seifert.