Another technique: thought quarantine. New ideas should have to endure an observation period and careful testing before they enter your repertoire, no matter how convincing they seem. If you adopt them too quickly, you risk becoming attached to them before you have a chance to notice their flaws.
I’ve found this to be particularly important with Eliezer’s posts on OB and here. Robin Hanson’s and most other posts are straightforward: they present data and then an interpretation. Eliezer’s writing style also communicates what it feels like to believe his interpretation. The result is that after reading one of Eliezer’s posts, my mind acts as though I believed what he was saying during the time I spend reading it. If I don’t suspend judgment on Eliezer’s ideas for a day or two while I consider them and their counter-arguments, they make themselves at home in my mind with inappropriate ease.
I won’t go so far as to accuse Eliezer of practicing the dark arts—I agree that communicating experiences is worthwhile. I read OB for the quality of the prose as well as idea content, both of which stand out among blogs. But while this effect hasn’t been articulated in comments as far as I know, I suspect that it contributes to the objections Robin Hanson and others have to Eliezer’s style.
I have to say, that goes a bit beyond what I intended. But that part where I communicate the experience is really important. I wonder if there’s some way to make it a bit less darkish without losing the experiential communication?
It may be time for a good Style vs Content Debate; first commenter to scream false dilemma gets a prize
It would be terribly dissapointing to see effective communication subordinated to efforts to avoid being labelled dark. The communication of experience serves as an effective reminder that we are reading not an infalible surce of wisdom, but just some guy who writes blog posts.
I’d really prefer people spent effort trying not to give me biassed information than spending that effort trying to persuade me that they aren’t dark.
To work on refining and verifying a theory, one first has to understand it, to be able to formulate theoretical conclusions given possible future data or arguments. While the validity of a theory in the current form isn’t certain, moving towards resolution of this uncertainty requires mastery of the theory, and it’s where the communication of experience comes in.
The direct analogy is with categorization: to measure the precision of a categorization technique, one first may need to implement the categorization algorithm. Resulting categories computed by the algorithm don’t claim to be the truth, they are merely labels with aspirations for predictive power.