{"id":23322,"date":"2026-04-30T17:34:11","date_gmt":"2026-04-30T17:34:11","guid":{"rendered":"https:\/\/www.europesays.com\/ai\/23322\/"},"modified":"2026-04-30T17:34:11","modified_gmt":"2026-04-30T17:34:11","slug":"chatgpt-is-going-to-stop-talking-about-goblins-and-gremlins-as-it-shedsnerdy-persona","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/ai\/23322\/","title":{"rendered":"ChatGPT is going to stop talking about goblins and gremlins as it sheds&#8221;nerdy&#8221; persona"},"content":{"rendered":"<p>If you\u2019ve been chatting with <a href=\"https:\/\/www.digitaltrends.com\/topic\/chatgpt\/\" rel=\"nofollow noopener\" target=\"_blank\">ChatGPT<\/a> lately and noticed it dropping oddly specific references to goblins, gremlins, ogres, or trolls, you\u2019re not imagining things. OpenAI has now <a href=\"https:\/\/openai.com\/index\/where-the-goblins-came-from\/\" rel=\"noopener noreferrer nofollow\" target=\"_blank\">explained<\/a> why ChatGPT has developed this strange habit and how it\u2019s getting fixed.<\/p>\n<p>How a \u201cnerdy\u201d quirk became everyone\u2019s problem<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" width=\"2104\" height=\"768\" data-wp-class--hide=\"state.isContentHidden\" data-wp-class--show=\"state.isContentVisible\" data-wp-init=\"callbacks.setButtonStyles\" data-wp-on-async--click=\"actions.showLightbox\" data-wp-on-async--load=\"callbacks.setButtonStyles\" data-wp-on-async-window--resize=\"callbacks.setButtonStyles\" src=\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/04\/ChatGPT-goblin-reference-increase.jpg\" alt=\"ChatGPT-goblin-reference-increase\" class=\"wp-image-5970004\"  \/><\/p>\n<p>\t\tOpenAI<\/p>\n<p>The problem quietly <a href=\"https:\/\/www.digitaltrends.com\/computing\/openai-launches-chatgpt-5-1-with-8-new-personalities-warmer-tone-and-quicker-replies\/\" rel=\"nofollow noopener\" target=\"_blank\">started with GPT-5.1, released in November<\/a>. After that launch, use of the word \u201cgoblin\u201d in ChatGPT responses jumped 175%, while \u201cgremlin\u201d rose 52%. The culprit turned out to be one of <a href=\"https:\/\/www.digitaltrends.com\/cool-tech\/chatgpt-now-lets-you-dial-up-the-warmth-or-tone-down-the-enthusiasm-in-its-responses\/\" rel=\"nofollow noopener\" target=\"_blank\">ChatGPT\u2019s optional personality settings<\/a> called \u201cNerdy,\u201d which was designed to make the AI sound playful and intellectually curious.<\/p>\n<p>During training, OpenAI accidentally gave the model unusually high rewards for responses that included creature-based metaphors, and the habit took hold fast.<\/p>\n<p>How did a single personality setting cause so much goblin talk?<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" width=\"2166\" height=\"1023\" data-wp-class--hide=\"state.isContentHidden\" data-wp-class--show=\"state.isContentVisible\" data-wp-init=\"callbacks.setButtonStyles\" data-wp-on-async--click=\"actions.showLightbox\" data-wp-on-async--load=\"callbacks.setButtonStyles\" data-wp-on-async-window--resize=\"callbacks.setButtonStyles\" src=\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/04\/ChatGPT-goblin-reference-2.jpeg\" alt=\"ChatGPT-goblin-reference\" class=\"wp-image-5970005\"  \/><\/p>\n<p>\t\tOpenAI<\/p>\n<p>Here\u2019s where it gets interesting. Even users who never switched on the Nerdy personality started seeing goblin references pop up in their chats. That\u2019s because AI training isn\u2019t contained to one setting. Once the <a data-popup-added=\"true\" style=\"cursor: pointer\" href=\"https:\/\/www.digitaltrends.com\/computing\/openai-explains-how-to-use-different-chatgpt-models\/\" rel=\"nofollow noopener\" target=\"_blank\">ChatGPT model<\/a> was rewarded for that style, the behavior bled into general responses across the board.<\/p>\n<p>OpenAI says the Nerdy personality made up just 2.5% of all ChatGPT responses, yet accounted for 66.7% of all goblin mentions.<\/p>\n<p>So how is OpenAI actually fixing this?<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" width=\"2000\" height=\"1200\" data-wp-class--hide=\"state.isContentHidden\" data-wp-class--show=\"state.isContentVisible\" data-wp-init=\"callbacks.setButtonStyles\" data-wp-on-async--click=\"actions.showLightbox\" data-wp-on-async--load=\"callbacks.setButtonStyles\" data-wp-on-async-window--resize=\"callbacks.setButtonStyles\" src=\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/04\/chatgpt5.1-launched-with-8-personalities.jpg\" alt=\"chatgpt5.1-launched-with-8-personalities\" class=\"wp-image-5762984\"\/><\/p>\n<p>\t\tOpenAI<\/p>\n<p>OpenAI retired the Nerdy personality in March with <a data-popup-added=\"true\" style=\"cursor: pointer\" href=\"https:\/\/www.digitaltrends.com\/computing\/openais-new-gpt-5-4-model-makes-chatgpt-better-at-handling-your-complex-multi-step-workflows\/\" rel=\"nofollow noopener\" target=\"_blank\">ChatGPT-5.4<\/a>, which caused goblin references to drop sharply. The company also stripped out the reward signal driving the behavior and filtered training data to reduce references to other magical creatures.<\/p>\n<p><a href=\"https:\/\/www.digitaltrends.com\/computing\/if-youre-into-ai-coding-openai-just-puts-its-codex-on-windows\/\" rel=\"nofollow noopener\" target=\"_blank\">Its coding tool, Codex<\/a>, however, needed a separate override instruction since it had already begun training before the root cause was identified. Fantasy fans can still unlock goblin mode in Codex manually, if that\u2019s your thing.<\/p>\n<p>OpenAI is also dealing with other personality-related decisions, including putting its previously teased <a href=\"https:\/\/www.digitaltrends.com\/computing\/chatgpt-is-not-getting-an-erotic-mode-after-all\/\" rel=\"nofollow noopener\" target=\"_blank\">adult mode for verified users on hold indefinitely<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"If you\u2019ve been chatting with ChatGPT lately and noticed it dropping oddly specific references to goblins, gremlins, ogres,&hellip;\n","protected":false},"author":2,"featured_media":23323,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[7],"tags":[24,1604,25,580,15993,1221,15994,15995,157],"class_list":{"0":"post-23322","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-openai","8":"tag-ai","9":"tag-ai-chatbot","10":"tag-artificial-intelligence","11":"tag-chatgpt","12":"tag-chatgpt-personality","13":"tag-computing","14":"tag-goblins","15":"tag-gremlins","16":"tag-openai"},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/23322","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/comments?post=23322"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/23322\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media\/23323"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media?parent=23322"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/categories?post=23322"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/tags?post=23322"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}