Some thoughts on how to turn conversations about social-reality, into conversations about reality:
The main skill I use (which I still need to practice more) is to find something about the discussion that I’m honestly uncertain or confused about—somewhere that my models stop—and get curious
My description of this looks like: Start proposing mechanisms that would predict the phenomena, notice which parts of the mechanism your System 1 feels iffy about, and attempt to improve/modify the mechanism. Then iterate.
A high-effort way to go straight to this skill is to run a simulation of a person you know who is endlessly curious about the object level details of the world. That person who is always asking the physics/economics/astronomy/etc questions. Simulate them in your situation and see what they’d start talking about.
A low-effort way to imitate this skill (to help you bootstrap up) might be to pick some part of the conversational topic, and go through the process of a 5 Whys analysis. Practising the process can help your System 1 learn the purpose of the process.
(Jacobjacob suggested 5 Whys to me in person, inspiring me to write this whole comment)
Another vote for ideas on how to steer away from sticky, low-productivity conversations.
Some thoughts on how to turn conversations about social-reality, into conversations about reality:
The main skill I use (which I still need to practice more) is to find something about the discussion that I’m honestly uncertain or confused about—somewhere that my models stop—and get curious
My description of this looks like: Start proposing mechanisms that would predict the phenomena, notice which parts of the mechanism your System 1 feels iffy about, and attempt to improve/modify the mechanism. Then iterate.
A high-effort way to go straight to this skill is to run a simulation of a person you know who is endlessly curious about the object level details of the world. That person who is always asking the physics/economics/astronomy/etc questions. Simulate them in your situation and see what they’d start talking about.
A low-effort way to imitate this skill (to help you bootstrap up) might be to pick some part of the conversational topic, and go through the process of a 5 Whys analysis. Practising the process can help your System 1 learn the purpose of the process.
(Jacobjacob suggested 5 Whys to me in person, inspiring me to write this whole comment)