Promoted to curated: I think the question of “what makes a good explanation, and how do humans come to really understand things?” is one of the core questions of rationality. I think this post is a well-written and clear attempt at introducing some important considerations on what makes a good explanation, and I expect most readers to walk away with a slightly improved ability to give better explanations than they were before.
Importantly, in the broader idea-pipeline of LessWrong, I think the concept outlined in this post is still in a relatively early poetry phase, and I would be somewhat hesitant for it to be adopted widely. I think as we develop and analyze the ideas in the post further, I expect we will eventually get something more similar to Eliezer’s “A technical explanation of a technical explanation”, where we can be more precise and robust in specifying what makes a good explanation, instead of having to rely on vaguer metaphors and individual examples.
(I don’t mean to say that this post says the same thing as Eliezer’s technical explanation post. I think it primarily talks about different aspects, that are also important. I am only trying to say that Eliezer’s technical explanation seems like a good target standard for rigor and robustness)
Promoted to curated: I think the question of “what makes a good explanation, and how do humans come to really understand things?” is one of the core questions of rationality. I think this post is a well-written and clear attempt at introducing some important considerations on what makes a good explanation, and I expect most readers to walk away with a slightly improved ability to give better explanations than they were before.
Importantly, in the broader idea-pipeline of LessWrong, I think the concept outlined in this post is still in a relatively early poetry phase, and I would be somewhat hesitant for it to be adopted widely. I think as we develop and analyze the ideas in the post further, I expect we will eventually get something more similar to Eliezer’s “A technical explanation of a technical explanation”, where we can be more precise and robust in specifying what makes a good explanation, instead of having to rely on vaguer metaphors and individual examples.
(I don’t mean to say that this post says the same thing as Eliezer’s technical explanation post. I think it primarily talks about different aspects, that are also important. I am only trying to say that Eliezer’s technical explanation seems like a good target standard for rigor and robustness)
I agree on the limits of this post—I hope it’s a beginning, not an end.