On rationality skills

If you have a developed system for finding the difference between truth and falsehood, and constantly think about rationality skills, and are always improving them, learning new techniques, and keep thinking about meta-level things: congratulations, your skills are useless.

Well, they’re not entirely useless. There’s a theoretical world in which, were you to apply them, they would suddenly become useful.

There’s the notion that by doing meta-level work, you’re sort of recursively improving your ability to reason; by not pursuing to know the truth for an object-level problem, but instead by pursuing to know how to know truth, you’re making yourself more truth-seekingly powerful, you’re turning yourself into a Bene Gesserit, you’re becoming an instrument of divine truth-finding, a being capable of discerning truth among a thousand lies… Only that never happens.

You keep writing stupid little essays like this that are about the meta-level, about “being a rationalist”, and what a rationalist should do, instead of writing essays about things that have happened, or are happening, or will happen. (Or should happen.)

Does practicing chess make you smarter? Which books should you read if you want to become a successful warlord? What do telomeres actually do? What are the variables that determine if an immigration program is successful? What’s the moral case for eating shellfish? How likely is it that random ARM instructions cause a NOP sled? How to decentralize overcrowded cities and create smaller regional urban centers? Etc, etc, etc.

Yes, this whole essay in fact suffers from the very problem I’m trying to convey; that it’s close to useless to meta meta meta everything if you never do the actual thing. This is a letter to myself: learn the thing, do the thing, write about thing.