When I spell things out so clearly only an idiot could misinterpret my words, the comments get worse because idiots misinterpret my words.
Do you judge how successful a piece is by the quality of the comments it received? That strikes me as a strange metric to optimize for.
I could see this metric aligning with “discover”, if you permit bad comments to waste your time. But I think you get more “explain” by writing for a broad audience. Maybe not “dumb”, but at least “ignorant”.
Of course, comment quality is an input into your overall feedback. But not the only input, and importantly not the main one, I think. (By “comment” here I’m thinking of “random internet strangers saying things about your article, eg here on Lesswrong.”)
There are so many other sources of feedback, including:
how you yourself judge the article
feedback from people you trust
reshares, link backs, quotes of the article
up votes, views
Which I think should combine for a holistic evaluation of how well your particular article was received. Comments may be one of the easier metrics, but leaning on it too heavily runs afoul of “drunk looking under streetlight for keys”
At the risk of being a misinterpreting idiot:
Do you judge how successful a piece is by the quality of the comments it received? That strikes me as a strange metric to optimize for.
I could see this metric aligning with “discover”, if you permit bad comments to waste your time. But I think you get more “explain” by writing for a broad audience. Maybe not “dumb”, but at least “ignorant”.
Even math PHDs prefer simple English before differential equations. See Eliezer on “aim high, shoot low”: https://www.lesswrong.com/posts/2TPph4EGZ6trEbtku/explainers-shoot-high-aim-low
How could you not use comment quality as feedback? Often the alternative is no feedback at all
Of course, comment quality is an input into your overall feedback. But not the only input, and importantly not the main one, I think. (By “comment” here I’m thinking of “random internet strangers saying things about your article, eg here on Lesswrong.”)
There are so many other sources of feedback, including:
how you yourself judge the article
feedback from people you trust
reshares, link backs, quotes of the article
up votes, views
Which I think should combine for a holistic evaluation of how well your particular article was received. Comments may be one of the easier metrics, but leaning on it too heavily runs afoul of “drunk looking under streetlight for keys”