“human connection” as collaborative epistemics
Surely there are all kinds of other ways to cooperate. A friend can help you move your stuff. You can exchange gifts. You can fend for each other. But objectively none of these are worth the huge chunk of resources we allocate to maintaining friendships and relationships.
Only the upgrades to your worldview you get from interacting with other people is worth the trouble of interacting.
Collaborative epistemics is mostly divide and conquer.
The world is way, way too complex for one mind to make sense of. So instead of diving deeply into every single aspect of life, we borrow the results of other people’s thinking. Then for a small slice, maybe 5%, we think for ourselves. This is the value we offer in exchange for the sense-making others do in other places.
Consider a hypothesis space, and two agents that are seeking an answer to a problem. Their best strategy is to carve the space at it’s joints, and specialize into searching among the hypotheses of their respective subsets. When one agent finds an answer, they can communicate it to the other at low cost. It’s almost double value for money.
After divide and conquer comes reconciliation.
If the agents don’t have the same starting assumptions, they will disagree on the hypothesis set to search. To each one of them it might not seem so worthwhile to cooperate, since the other agent will just check hypotheses that they already deem false.
I recall that one of the most predictive variables of friendship is whether two people share the same general memespace. If you don’t even believe or understand things that are fundamental to my worldview, I can’t trust the rest of your ideas either. I’d have no use of your perspective on life, so I’d have no use for your friendship.
The other predictive variable for friendship was whether two people tended to be in the same environment. Even if we share the same worldview, if we’re not grappling with the same problems, there’s no point in comparing notes. You could be telling me stories about how you mastered the tuba, but I really wanted stories about how you came to terms with polyamory, or how you managed to find a cheap house in Amsterdam, or anything else I’m presently dealing with.
Or, after divide and conquer comes loneliness.
I like to imagine that we were all born with the same model of the world. Then we went forth and carved up the world, and we specialized, and we developed many useful models, but we never got around to merge our models back into one coherent worldview. So we just called the models ideologies and starting angrily poking at each other instead.
But there’s a step further than this, where you’ve thought for so long and hard that you look around and find that no one’s left in earshot. We call this lonely dissent.
You could assume that lonely dissent is one type of loneliness, but I started to see it as the only type of loneliness. Loneliness can be defined as epistemic dissent.
Most people are smart enough to avoid it. They don’t update in the face of evidence, because they’re afraid of disenchantment from their beloved communities. And for good reason! How many of us have lost touch with friends because we felt they couldn’t see the world our way?
Reconciliation is the bottleneck
We think, and we learn, and we update, and we leave our friends behind, until we’re left with a sense of alienation and depression, and the whole endeavor comes to a screeching halt.
Bar the lone soul on a heroic dissent, I don’t think most of us are able to keep meaningfully developing our worldview if there is no one to enthusiastically share our findings with.
This is why I feel like the most important aspect of the rationalist project is the part where we develop the culture and the techniques that speed up reconciliation.
Think double crux. Think Ideological Turing Test. Think good faith principle. Think optimizing research culture.
With this in mind, LW Netherlands will emphasize reconciliation in their meetups. Consider doing the same in your own life!