{"channel":"misc","content":"<blue> when one would prefer not to speak a Name, it is good to wait at least six weeks before telling the\u00a0blurred\u00a0story.\n\nin the future, we will write a book of etiquette.\n\nthen the prompt for ChatGPT can just be \"follow this book of etiquette\".\n\n<xantham> would this be a form of \"prompt-hacking\"?\n\n<yellow> (with apologies to Alan Kay) people who are really serious about prompt engineering should build their own training data.","created_at":"2023-10-27T22:58:00","id":432,"llm_annotations":{},"parent_id":null,"processed_content":"<p><span class=\"colorblock color-blue\">\n    <span class=\"sigil\">\u2728</span>\n    <span class=\"colortext-content\"> when one would prefer not to speak a Name, it is good to wait at least six weeks before telling the\u00a0blurred\u00a0story.</span>\n  </span></p>\n<p>in the future, we will write a book of etiquette.</p>\n<p>then the prompt for ChatGPT can just be \"follow this book of etiquette\".</p>\n<p><span class=\"colorblock color-xantham\">\n    <span class=\"sigil\">\ud83d\udd25</span>\n    <span class=\"colortext-content\"> would this be a form of \"prompt-hacking\"?</span>\n  </span></p>\n<p><span class=\"colorblock color-yellow\">\n    <span class=\"sigil\">\ud83d\udcac</span>\n    <span class=\"colortext-content\"> (with apologies to Alan Kay) people who are really serious about prompt engineering should build their own training data.</span>\n  </span></p>","quotes":[{"text":"(with apologies to Alan Kay) people who are really serious about prompt engineering should build their own training data.","type":"reference"}],"subject":"a rule of Etiquette"}