{"id":13888,"date":"2024-06-24T09:16:27","date_gmt":"2024-06-24T01:16:27","guid":{"rendered":"https:\/\/www.1ai.net\/?p=13888"},"modified":"2024-06-24T09:16:27","modified_gmt":"2024-06-24T01:16:27","slug":"%e7%a0%94%e7%a9%b6%e5%8f%91%e7%8e%b0%ef%bc%8copenai-%e7%9a%84-gpt-4o-%e9%81%93%e5%be%b7%e6%8e%a8%e7%90%86%e8%83%bd%e5%8a%9b%e8%83%9c%e8%bf%87%e4%ba%ba%e7%b1%bb%e4%b8%93%e5%ae%b6","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/13888.html","title":{"rendered":"OpenAI's GPT-4o outperforms human experts in moral reasoning, study finds"},"content":{"rendered":"

A recent study showed thatOpenAI<\/a> Latest Chatbots GPT-4o<\/a> Ability to provide ethical explanations and advice of a higher quality than that provided by \u201crecognized\u201d ethical experts.<\/p>\n

According to The Decoder on Saturday local time, researchers from the University of North Carolina at Chapel Hill and the Allen Institute for AI conducted two studies to compare the GPT model with human moral reasoning ability to explore whether large language models can be considered "Ethics Expert<\/strong>\u201d.<\/p>\n

The research contents are summarized as follows:<\/p>\n

Study 1<\/h3>\n

501 American adults compared\u00a0GPT-3.5-turbo model<\/strong>and other human participants\u2019 moral explanations. The results showed that people rated GPT\u2019s explanations as more ethical than human participants\u2019 explanations.More ethical, more trustworthy, more thoughtful<\/strong>.<\/p>\n

Evaluators also rated AI-generated evaluations as more reliable than others. While the differences were small, the key finding is that AI canMatch or even surpass<\/strong>Human-level moral reasoning.<\/p>\n

\"6838d680-3d18-4109-a83e-5676818e9e77\"<\/p>\n

Study 2<\/h3>\n

The advice generated by OpenAI\u2019s latest GPT-4o model was compared to that of renowned ethicist Kwame Anthony Appiah in The Ethicist column in The New York Times. 900 participants rated the quality of the advice on 50 \u201cethical dilemmas.\u201d<\/p>\n

The results show that GPT-4o isAlmost every aspect<\/strong>\u201d outperformed human experts. People believed that AI-generated recommendationsMore morally correct, more trustworthy, more thoughtful, more accurate<\/strong>Only in terms of perceiving nuances was there no significant difference between the AI and human experts.<\/p>\n

\"db641aab-c4cb-456a-9fb5-0d52b4591e35\"<\/p>\n

The researchers believe these results suggest that AI can pass the "Comparative Moral Turing Test" (cMTT). And text analysis shows that GPT-4o usesMoral and Positive Language<\/strong>More than human experts. This may partly explain why the AI\u2019s recommendations were rated higher \u2014 but it\u2019s not the only factor.<\/p>\n

It\u2019s important to note that this study was conducted only on American participants, and further research is needed to explore cultural differences in how people view AI-generated moral reasoning.<\/p>\n

Paper address:<\/strong>https:\/\/osf.io\/preprints\/psyarxiv\/w7236<\/span><\/a><\/p>","protected":false},"excerpt":{"rendered":"

A recent study has shown that OpenAI's newest chatbot, GPT-4o, is able to provide ethical explanations and advice that are better than those provided by \"recognized\" ethical experts. Researchers at the University of North Carolina at Chapel Hill and the Allen Institute for AI Research conducted two studies comparing the GPT model to human moral reasoning to see if the big language model could be considered a \"moral expert,\" The Decoder reported Saturday. A summary of the studies is as follows: Study 1 501 American adults compared the GPT-3.5-turbo model with the moral interpretations of other human participants. The results indicated that the GPT's explanations were perceived to be more ethical than those of the human participants.<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[2582,190],"collection":[],"class_list":["post-13888","post","type-post","status-publish","format-standard","hentry","category-news","tag-gpt-4o","tag-openai"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/13888","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=13888"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/13888\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=13888"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=13888"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=13888"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=13888"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}