x
Reducing x-risk might be actively harmful — LessWrong