This is a special post for quick takes by Evie. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
Mentioned in
Sorted by Click to highlight new quick takes since: Today at 5:22 AM

A distinction I've found useful is "object-level" vs "social reality". They are both adjectives that describe types of conversation/ ideas.

Object-level discussions are about ideas and actions (e.g. AI timelines, the mechanics of launching a successful startup). Object-level ideas are technical, empirical, and often testable. Object-level refers to what ideas are important or make sense. It is focused on truth-seeking and presenting arguments clearly. 

Social reality discussions are about people and organisations (e.g. Will MacAskill, Open Philanthropy). Social reality is more meta, more abstract, and less testable than object-level. Social reality refers to which people are influential/powerful (and what they think), how to network with people, how to persuade people.

Object-level: What's the probability of extinction this century?

Social reality: What does Toby Ord think is the probability of extinction this century?

I have found it very helpful to start labelling whether I'm in object-level conversation mode vs social reality conversation mode. It helps me notice when I'm deferring without having thought about it (e.g. "well, Will MacAskill says [x]"  instead of asking myself what I think about [x]), or when I fall into a mode of chit-chatting about the who's-who of EA, instead of trying to truth-seek (of course, chit chatting sometimes is fine -- I just want to be intentional about when I'm doing it).

And social reality isn't necessarily bad, but it's helpful to flag when a conversation enters "social reality mode."

I do think it's good for many/more/most conversations to centre around the object-level. I am personally trying to move my ratio more towards object-level.

(This was a core theme of an Atlas camp I attended, which I found extremely valuable. The above definitions are loosely based on a message from Jonas, but I didn't run them by him before posting.)

This relates to a caveat in my recent post:

  • I’m concerned about too much EA meta conversation — about worlds where most of EA dialogue is talking about EAs talking about EAs (lots of social reality and not enough object-level). 
  • These sorts of convos are often very far removed from {concrete things that help the world}, and I worry about them taking away attention from more important ones. 
  • I think it’s probably much better (for the world) for conversations to stay focused on the real world, object-level claims and arguments.

Part of me wants to flesh this thought out properly soon. But even this conversation is meta! And I'm trying to encourage/ focus more on object-level ideas. So do I write it? I'm not sure. 

[anonymous]1y0
0
1

I feel like defining a separate term for this may be better. Maybe:

Integrated view = defensible view + delegated view

where defensible view contains details you personally can defend, and delegated view includes other people's views and whatever factor of delegation you feel is appropriate