x
Some comments on the CAIS paradigm — LessWrong