For this Sunday's LessWrong meetup, I'll be giving a short presentation on some moral/game-theoretic/coordination problems and solutions I've been thinking about, which I would like feedback on.
Three related puzzles:
- Sometimes, people disagree on coordination protocols, or on moral frameworks. What do we do about that in the general case?
- Many rationalists explore novel coordination principles. This can result in us learning about "new and exciting ways that everyone is defecting on each other all the time". If a principle is novel, it's hard to coordinate around. How do we handle disagreements about that?
- Sometimes, Alice and Bob disagree on how to coordinate, but Bob thinks Alice has overall demonstrated better judgment, and it's (relatively) easy for Bob to defer to Alice. But other times, Alice and Bob don't trust each other, and each of them thinks the other is somewhat less sophisticated (despite having overall similar worldviews). How should they handle that situation?
"The Coordination Frontier" is my (placeholder) term for "the parts of morality/game-theory/coordination that aren't obvious, especially new principles that are novel advances on the current state-of-the-art." I think it's a useful concept for us to collectively have as we navigate complex new domains in the coming years.
I have some existing thoughts that I'd like feedback on, and I'd generally like to spark discussion about this topic.
Approximate format will be:
- I give a short(ish) presentation
- General Q&A and public discussion
- Brainstorming in a google doc, and then splitting into smaller groups to discuss particular subtopics.
- Return to a centralized conversation and share insights.
This will take place in the Walled Garden, in the Manor (to the north of the central map)