x
Aligned AI Proposals — LessWrong