One possible reason for favoring the wiki, at least in the long run: a well-structured wiki is more discoverable and accessible. The huge branching tree of your OB posts is daunting enough; the entire past corpus of LW is rapidly going to become even more intimidating for lacking a single author with consistent style and themes. Imagine LW at roughly the same rate of content generation and a slowly expanding user base, three years from now. Where would someone even start?
Distilling the ideas into a manageable synthesis of what’s been discussed in multiple posts, with extensive inter-linking, terms defined, and links to relevant posts would be far more useful for a newcomer than just minimal definitions with links to posts.
To use an academic analogy, if LW posts are journal articles, the wiki ought to be a textbook.
One possible reason for favoring the wiki, at least in the long run: a well-structured wiki is more discoverable and accessible. The huge branching tree of your OB posts is daunting enough; the entire past corpus of LW is rapidly going to become even more intimidating for lacking a single author with consistent style and themes. Imagine LW at roughly the same rate of content generation and a slowly expanding user base, three years from now. Where would someone even start?
Distilling the ideas into a manageable synthesis of what’s been discussed in multiple posts, with extensive inter-linking, terms defined, and links to relevant posts would be far more useful for a newcomer than just minimal definitions with links to posts.
To use an academic analogy, if LW posts are journal articles, the wiki ought to be a textbook.
We just need better scholarship in the LW posts themselves, with review articles and tutorial articles, and possibly thematic workshops.