Response to questions from member Stoffer regarding the security risks of advanced artificial intelligence