Anti-akrasia remote monitoring experiment — LessWrong