The Conversations We Make Space For

By Severin @ 2022-07-28T21:36 (+22)

(Cross-posted to LessWrong)

While developing the necessary skills and models for making The Berlin Hub as great as it can be, I have had countless opportunities to talk and think about the norms, values, and rituals of the EA and rationalist communities. So here is a list of three types of conversations that we tend to make space for, sometimes in unique ways. I hope that making these explicit will help community builders reflect on which kinds of events to run, why to run them, and how.

1) Truth-seeking conversations

For rationalists, truth-seeking conversations are a way to cultivate epistemic rationality. For EAs, they are a crucial part of “using evidence and reason” to pursue their altruistic goals — according to CEA’s definition of effective altruism as “using evidence and reason to figure out how to benefit others as much as possible, and taking action on that basis.”

We can divide truth-seeking conversations into three types:

1. Didactic: one person knows something the others don’t. The person with the knowledge teaches; the others listen, ask questions, and disagree if they think something is off.

Example: talks.

2. Symmetrical - assisting: one person listens to the other to help them think through an issue. Can happen with or without repeated targeted questions.

Examples: socratic ducking, socratic questioning, Street Epistemology.

3. Symmetrical - collaborative: two people disagree with one another about an issue. They sit down and talk things through in the spirit of collaborative truth-seeking.

Example: the archetypal method for collaborative truth-seeking is Double Crux. 

The key virtues of truth-seeking conversations are:

• to be aware of what you know and what not,
• to be curious,
• to admit when you are wrong,
• to share what you know, if and only if it is wanted.

2) Impact-focused conversations

For rationalists, impact-focused conversations are a way of cultivating instrumental rationality. For EAs, these conversations put the “figure out how to benefit others as much as possible” into “using evidence and reason to figure out how to benefit others as much as possible, and taking action on that basis.”

Once again, we can distinguish one asymmetrical and two symmetrical types:

1. Didactic: one person has crucial information about relevant parts of the world. They teach their mental models and give advice to others in the community.

Examples: talks, certain kinds of career coaching, informal conversations between community members with different levels and areas of expertise.

2. Symmetrical - assisting: one person (e.g., a coach) brings methodological expertise, the other one knowledge about their own situation. The former guides the latter through a process of figuring out where they are now, what they want to change, and which concrete next steps this change requires.

Examples: some types of life coaching, hamming circles on career-related issues.

3. Symmetrical - collaborative: all those involved bring an understanding both of the topic at hand and the process of how to get from here to the desired impact. The conversation flows freely as a collaborative exploration of possibilities.

Example: Step 1: drop a conversation starter like “What would you do if you had one billion dollars to fix all the world’s problems?” or “How could we entirely stop AI capabilities research?” into a room full of EAs and rationalists. Step 2: take notes, murphyjitsu, and follow through. Step 3: impact.

The key virtues of impact-focused conversation are:

• to have something to protect, and
• to become a person who actually does things.

3) Debugging conversations

For rationalists, these conversations are about actually getting things done instead of cargo-culting rationality (so to speak, they are about applied epistemic and instrumental rationality).

For EAs, they are about the “taking action on that basis”-part of “using evidence and reason to figure out how to benefit others as much as possible, and taking action on that basis.”

Debugging conversations are all about figuring out the problems in productivity, emotional life, relationships, etc. which hinder us from achieving what we could. Here, too, the threefold typology applies:

1. Didactic: one person has crucial skills and tools that can help with peoples’ productivity, mental health, etc. They teach these skills to those who want to acquire them.

Examples: productivity coaching, CFAR workshops.

2. Symmetrical - assisting: structurally similar to 2b), but with different goals.

Examples: some types of life coaching, hamming circles on personal issues.

3. Symmetrical - collaborative: two or more people sit down to figure out a certain issue which neither of them feel like they have a complete grasp of.

Examples: discussion rounds on impostor syndrome or productivity-related issues.

The key virtues of debugging conversations are:

• tsuyoku naritai (to want to become stronger),
• to become a person who actually does things, and
• to practice self-compassion along the way.

Which types of conversations happen in the spaces you are involved in? Which ones do you want to see more? What can others learn from your ways of making them happen?