x
Rafael Kaufmann: Gaia: Distributed planetary-scale AI safety — LessWrong