Kaj_Sotala comments on Three new papers on AI risk