Over the years I’ve thought about a “LessWrong/Alignment” journal article format the way regular papers have Abstract-Intro-Methods-Results-Discussion. Something like that, but tailored to our needs, maybe also bringing in OpenPhil-style reasoning transparency (but doing a better job of communicating models).
Such a format could possibly mandate what you’re wanting here.
I think it’s tricky. You have to believe any such format actually makes posts better rather than constraining them, and it’s worth the effort of writers to confirm.
It is something I’d like to experiment with though.
Over the years I’ve thought about a “LessWrong/Alignment” journal article format the way regular papers have Abstract-Intro-Methods-Results-Discussion. Something like that, but tailored to our needs, maybe also bringing in OpenPhil-style reasoning transparency (but doing a better job of communicating models).
Such a format could possibly mandate what you’re wanting here.
I think it’s tricky. You have to believe any such format actually makes posts better rather than constraining them, and it’s worth the effort of writers to confirm.
It is something I’d like to experiment with though.