(Meta: I’m not sure if the following comment is valuable. What you wrote really resonated, though.)
I’ve felt for some time that, despite being “part of the community” for ( … checks … ) over seven years, I’m at around a yellow belt at rationality. Maybe an honorary green belt. I know a lot about rationality. I can talk a good game. Somebody can watch a lot of UFC and learn to rattle off the names of the techniques without being able to do any of them.
I use the belts metaphor deliberately. I think you’re right that “black belt rationality” is composed of a lot of gears. A white belt starts with simple movements and refines control, perception and understanding. They do this by following a prescribed path of development. The various subskills are nurtured along the way and culminate in a superior warrior. I’m not sure what that path looks like for rationality. I’m not sure what the paths are, what the gears are, or in what order they need to be developed. I could sketch something out, with all the right buzzwords. But that map would be as likely to be optimal as a plan for martial arts skill sketched out by an average UFC fan.
At the risk of being a bit cringey: I feel, when reading Eliezer’s writing, the same what the I feel watching Anderson Silva in a fight. I realize that I’m witnessing mastery of something. But Eliezer seems to be something like a “natural athlete” of rationality. Many highly successful people exhibit specific refined rationality-gears by age 16. Gears that most people grow old and die without ever discovering. Studying Eliezer doesn’t really tell me what to do to improve. Even Eliezer’s own advice might not be worth much. He may not realize that his students may completely lack gears that he had polished by puberty.
(I’m focusing on Eliezer here because he was the guy who made the first, boldest, and most successful stab at being Rationality Sensei. And I’m not criticizing what he accomplished, I’m reflecting on why I don’t feel like I’m as far along as I could be.)
I’ve wished for a long time that there was a real belt system for this stuff. I wrote all this because I have a similar sense that I’m progressing down a path that I’m forging as I go. Yet there are clearly other people who are further along the path. Try as I might, I can’t reach their level by aping their movements.
If you want to repeat Eliezer’s path to success, I’m pretty sure the limiting factor is creative writing, not rationality. But maybe creative writing isn’t even fun for you. It’s better to find your own path.
I think part of the thing is that, although there are some clearly defined skills and levels and whatnot, we’re still very much in the process of learning a) what skills are most important, b) how to teach them (in some cases, how to teach them at all)
There’s certain things Eliezer is good at, and some very different things that, say, Nate Soares is good at that Eliezer self-identifies as not-very-good-at, and different-still things that Luke Mulhauser is good at, and a host of other people, each good at different things that seem clearly part of the rationality canon but which go off in different directions. The concept of “Belt” feels too linear to encapsulate them.
My sense is that CFAR represents the body-of-knowledge of “what we have some sense of how to teach” to arbitrary people, which (I don’t think?) yet includes much of the higher-level stuff.
(Meta: I’m not sure if the following comment is valuable. What you wrote really resonated, though.)
I’ve felt for some time that, despite being “part of the community” for ( … checks … ) over seven years, I’m at around a yellow belt at rationality. Maybe an honorary green belt. I know a lot about rationality. I can talk a good game. Somebody can watch a lot of UFC and learn to rattle off the names of the techniques without being able to do any of them.
I use the belts metaphor deliberately. I think you’re right that “black belt rationality” is composed of a lot of gears. A white belt starts with simple movements and refines control, perception and understanding. They do this by following a prescribed path of development. The various subskills are nurtured along the way and culminate in a superior warrior. I’m not sure what that path looks like for rationality. I’m not sure what the paths are, what the gears are, or in what order they need to be developed. I could sketch something out, with all the right buzzwords. But that map would be as likely to be optimal as a plan for martial arts skill sketched out by an average UFC fan.
At the risk of being a bit cringey: I feel, when reading Eliezer’s writing, the same what the I feel watching Anderson Silva in a fight. I realize that I’m witnessing mastery of something. But Eliezer seems to be something like a “natural athlete” of rationality. Many highly successful people exhibit specific refined rationality-gears by age 16. Gears that most people grow old and die without ever discovering. Studying Eliezer doesn’t really tell me what to do to improve. Even Eliezer’s own advice might not be worth much. He may not realize that his students may completely lack gears that he had polished by puberty.
(I’m focusing on Eliezer here because he was the guy who made the first, boldest, and most successful stab at being Rationality Sensei. And I’m not criticizing what he accomplished, I’m reflecting on why I don’t feel like I’m as far along as I could be.)
I’ve wished for a long time that there was a real belt system for this stuff. I wrote all this because I have a similar sense that I’m progressing down a path that I’m forging as I go. Yet there are clearly other people who are further along the path. Try as I might, I can’t reach their level by aping their movements.
If you want to repeat Eliezer’s path to success, I’m pretty sure the limiting factor is creative writing, not rationality. But maybe creative writing isn’t even fun for you. It’s better to find your own path.
I think part of the thing is that, although there are some clearly defined skills and levels and whatnot, we’re still very much in the process of learning a) what skills are most important, b) how to teach them (in some cases, how to teach them at all)
There’s certain things Eliezer is good at, and some very different things that, say, Nate Soares is good at that Eliezer self-identifies as not-very-good-at, and different-still things that Luke Mulhauser is good at, and a host of other people, each good at different things that seem clearly part of the rationality canon but which go off in different directions. The concept of “Belt” feels too linear to encapsulate them.
My sense is that CFAR represents the body-of-knowledge of “what we have some sense of how to teach” to arbitrary people, which (I don’t think?) yet includes much of the higher-level stuff.