{"id":863,"date":"2023-03-20T09:35:34","date_gmt":"2023-03-20T09:35:34","guid":{"rendered":"https:\/\/gyfted.blog\/?p=863"},"modified":"2025-05-22T13:42:02","modified_gmt":"2025-05-22T13:42:02","slug":"update-chatgpt-passes-at-adult-level-theory-of-mind-test","status":"publish","type":"post","link":"https:\/\/www.gyfted.me\/resources\/blog\/update-chatgpt-passes-at-adult-level-theory-of-mind-test","title":{"rendered":"ChatGPT passes human level Theory of Mind and other tests"},"content":{"rendered":"\n<p>Quick update to research on OpenAI&#8217;s ChatGPT-4 ability to pass standardized tests and empathy (theory of mind) assessment. In December&#8217;s research ChatGPT-3.5 could pass the below at a 9-year old&#8217;s level, you can <a href=\"https:\/\/gyfted.blog\/2023\/02\/20\/stanfords-kosinski-exciting-chatgpt-to-theory-of-mind-test\/\" target=\"_blank\" rel=\"noopener\">read more about it here<\/a>. <\/p>\n\n\n\n<p>Now, with GPT-4* ChatGPT can pass at an adult human level the <em>Faux-pas Recognition Test (Adult version)<\/em>, a validated test of Theory of Mind. <\/p>\n\n\n\n<p>See below: <a href=\"https:\/\/spectrum.ieee.org\/theory-of-mind-ai\" target=\"_blank\" rel=\"noopener\">https:\/\/spectrum.ieee.org\/theory-of-mind-ai<\/a> <\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"257\" src=\"https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-1024x257.png\" alt=\"\" class=\"wp-image-1628\" srcset=\"https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-1024x257.png 1024w, https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-300x75.png 300w, https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-768x193.png 768w, https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-1536x385.png 1536w, https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-2048x514.png 2048w, https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-1000x251.png 1000w, https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-230x58.png 230w, https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-350x88.png 350w, https:\/\/www.gyfted.me\/resources\/wp-content\/uploads\/2023\/03\/image-6-480x120.png 480w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>Wonder what&#8217;ll happen in image recognition form in reading emotions from people&#8217;s eyes &#8211; like in this <a href=\"https:\/\/www.gyfted.me\/quiz-landing\/social-intelligence\">emotion recognition assessment<\/a> we have created at Gyfted. We&#8217;re very sceptical of &#8216;reasoning&#8217; and symbolic system representation of any artificial intelligence algos, rather, they&#8217;re great at large scale computing and pattern recognition.. <\/p>\n\n\n\n<p><strong>2024 update<\/strong>: this is all really intriguing, this research is based on memory storage really, when we look at &#8211; for instance &#8211; <a href=\"https:\/\/machinelearning.apple.com\/research\/gsm-symbolic\" target=\"_blank\" rel=\"noopener\">Apple AI team&#8217;s research<\/a> and the fact that LLMs are able to perform extremely well in complex pattern matching, but <a href=\"https:\/\/arstechnica.com\/ai\/2024\/10\/llms-cant-perform-genuine-logical-reasoning-apple-researchers-suggest\/\" target=\"_blank\" rel=\"noopener\">not reasoning itself<\/a>. <\/p>\n\n\n\n<figure class=\"wp-block-embed is-type-rich is-provider-reddit wp-block-embed-reddit\"><div class=\"wp-block-embed__wrapper\">\n<blockquote class=\"reddit-embed-bq\" style=\"height:500px\" ><a href=\"https:\/\/www.reddit.com\/r\/OpenAI\/comments\/1g26o4b\/apple_research_paper_llms_cannot_reason_they_rely\/\" target=\"_blank\" rel=\"noopener\">Apple Research Paper : LLM\u2019s cannot reason . They rely on complex pattern matching . <\/a><br> by<a href=\"https:\/\/www.reddit.com\/user\/hasanahmad\/\" target=\"_blank\" rel=\"noopener\">u\/hasanahmad<\/a> in<a href=\"https:\/\/www.reddit.com\/r\/OpenAI\/\" target=\"_blank\" rel=\"noopener\">OpenAI<\/a><\/blockquote><script async src=\"https:\/\/embed.reddit.com\/widgets.js\" charset=\"UTF-8\"><\/script>\n<\/div><\/figure>\n","protected":false},"excerpt":{"rendered":"<p>Quick update to research on OpenAI&#8217;s ChatGPT-4 ability to pass standardized tests and empathy (theory of mind) assessment. In December&#8217;s research ChatGPT-3.5 [&hellip;]<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_eb_attr":"","postBodyCss":"","postBodyMargin":[],"postBodyPadding":[],"postBodyBackground":{"backgroundType":"classic","gradient":""},"enableOnePageScrollInPage":false,"onePageScrollSpeed":1000,"onePageScrollAnimation":"easeInOutQuad","onePageScrollShowDotNavigation":true,"onePageScrollNavigation":"scaleUp","onePageScrollNavigationPosition":"right","onePageScrollNavigationHorizontal":[],"onePageScrollNavigationVertical":[],"onePageScrollNavigationSpacing":[],"onePageScrollNavigationColor":"#00ff0d","onePageScrollNavigationColorHover":"#00ff0d","onePageScrollNavigationColorActive":"#00ff0d","onePageScrollNavigationIcon":[],"onePageScrollNavigationWidth":[],"onePageScrollNavigationWidthHover":[],"onePageScrollNavigationWidthActive":[],"onePageScrollNavigationHeight":[],"onePageScrollNavigationHeightHover":[],"onePageScrollNavigationHeightActive":[],"onePageScrollNavigationBorder":[],"onePageScrollNavigationBorderHover":[],"onePageScrollNavigationBorderActive":[],"onePageScrollNavigationBorderRadius":[],"onePageScrollNavigationBorderRadiusHover":[],"onePageScrollNavigationBorderRadiusActive":[],"onePageScrollNavigationTooltipTypography":[],"onePageScrollNavigationTooltipColor":"#ffffff","onePageScrollNavigationTooltipColorHover":"","onePageScrollNavigationTooltipBgColor":"#00ff0d","onePageScrollNavigationTooltipBgColorHover":"","onePageScrollNavigationTooltipPadding":[],"onePageScrollNavigationTooltipPaddingHover":[],"onePageScrollNavigationTooltipBorderRadius":[],"onePageScrollNavigationTooltipBorderRadiusHover":[],"pageSettingsCustomCss":"","_daim_seo_power":"1000","_daim_enable_ail":"1","footnotes":""},"categories":[36],"tags":[108,109,110,111],"class_list":["post-863","post","type-post","status-publish","format-standard","hentry","category-blog","tag-chatgpt","tag-openai","tag-testing","tag-theory-of-mind"],"_links":{"self":[{"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/posts\/863","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/comments?post=863"}],"version-history":[{"count":1,"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/posts\/863\/revisions"}],"predecessor-version":[{"id":1903,"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/posts\/863\/revisions\/1903"}],"wp:attachment":[{"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/media?parent=863"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/categories?post=863"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.gyfted.me\/resources\/wp-json\/wp\/v2\/tags?post=863"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}