{"id":22651,"date":"2023-09-13T04:13:46","date_gmt":"2023-09-13T02:13:46","guid":{"rendered":"https:\/\/www.wjst.de\/blog\/?p=22651"},"modified":"2023-09-14T19:03:05","modified_gmt":"2023-09-14T17:03:05","slug":"no-way-to-recognize-ai-text-blocks","status":"publish","type":"post","link":"https:\/\/www.wjst.de\/blog\/sciencesurf\/2023\/09\/no-way-to-recognize-ai-text-blocks\/","title":{"rendered":"No way to recognize AI generated text"},"content":{"rendered":"<p>Whatever I wrote before different methods to detect <a href=\"https:\/\/uk.pcmag.com\/help\/146686\/5-ways-to-detect-chatgpt-written-text\">AI written text<\/a> (using AI Text Classifer, GPTZero, Originality.AI&#8230;) seems now to be too optimistic. <a href=\"https:\/\/help.openai.com\/en\/articles\/8313351-how-can-educators-respond-to-students-presenting-ai-generated-content-as-their-own\">OpenAI even reports <\/a> that AI detectors do not work at all<\/p>\n<blockquote><p>While some (including OpenAI) have released tools that purport to detect AI-generated content, none of these have proven to reliably distinguish between AI-generated and human-generated content.<br \/>\nAdditionally, ChatGPT has no \u201cknowledge\u201d of what content could be AI-generated. It will sometimes make up responses to questions like \u201cdid you write this [essay]?\u201d or \u201ccould this have been written by AI?\u201d These responses are random and have no basis in fact.<br \/>\n&#8230;<br \/>\nWhen we at OpenAI tried to train an AI-generated content detector, we found that it labeled human-written text like Shakespeare and the Declaration of Independence as AI-generated.<br \/>\n&#8230;<br \/>\nEven if these tools could accurately identify AI-generated content (which they cannot yet), students can make small edits to evade detection.<\/p><\/blockquote>\n<p>BUT &#8211; according to a recent <a href=\"https:\/\/copyleaks.com\/blog\/copyleaks-ai-plagiarism-analysis-report\">Copyleaks<\/a> study, use of AI runs at high risk of plagiarizing earlier text that has been used to train the AI model. So it will be dangerous for everybody who is trying to cheat.<\/p>\n<figure id=\"attachment_22666\" aria-describedby=\"caption-attachment-22666\" style=\"width: 620px\" class=\"wp-caption alignnone\"><a href=\"https:\/\/www.wjst.de\/blog\/wp-content\/uploads\/2023\/09\/Bildschirmfoto-2023-09-14-um-18.52.59.jpg\" rel=\"key\" data-rel=\"key-image-0\" data-rl_title=\"\" data-rl_caption=\"\" title=\"\"><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-22666 size-medium\" src=\"https:\/\/www.wjst.de\/blog\/wp-content\/uploads\/2023\/09\/Bildschirmfoto-2023-09-14-um-18.52.59-620x602.jpg\" alt=\"\" width=\"620\" height=\"602\" srcset=\"https:\/\/www.wjst.de\/blog\/wp-content\/uploads\/2023\/09\/Bildschirmfoto-2023-09-14-um-18.52.59-620x602.jpg 620w, https:\/\/www.wjst.de\/blog\/wp-content\/uploads\/2023\/09\/Bildschirmfoto-2023-09-14-um-18.52.59-515x500.jpg 515w, https:\/\/www.wjst.de\/blog\/wp-content\/uploads\/2023\/09\/Bildschirmfoto-2023-09-14-um-18.52.59-768x746.jpg 768w, https:\/\/www.wjst.de\/blog\/wp-content\/uploads\/2023\/09\/Bildschirmfoto-2023-09-14-um-18.52.59.jpg 790w\" sizes=\"auto, (max-width: 620px) 100vw, 620px\" \/><\/a><figcaption id=\"caption-attachment-22666\" class=\"wp-caption-text\">https:\/\/copyleaks.com\/blog\/copyleaks-ai-plagiarism-analysis-report<\/figcaption><\/figure>\n\n<p>&nbsp;<\/p>\n<div class=\"bottom-note\">\n  <span class=\"mod1\">CC-BY-NC Science Surf , accessed 15.04.2026<\/span>\n <\/div>","protected":false},"excerpt":{"rendered":"<p>Whatever I wrote before different methods to detect AI written text (using AI Text Classifer, GPTZero, Originality.AI&#8230;) seems now to be too optimistic. OpenAI even reports that AI detectors do not work at all While some (including OpenAI) have released tools that purport to detect AI-generated content, none of these have proven to reliably distinguish &hellip; <a href=\"https:\/\/www.wjst.de\/blog\/sciencesurf\/2023\/09\/no-way-to-recognize-ai-text-blocks\/\" class=\"more-link\">Continue reading <span class=\"screen-reader-text\">No way to recognize AI generated text<\/span> <span class=\"meta-nav\">&rarr;<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5,9],"tags":[3358,4224,1841],"class_list":["post-22651","post","type-post","status-publish","format-standard","hentry","category-philosophy-of-science","category-computer-software","tag-ai","tag-detection","tag-plagiarism"],"_links":{"self":[{"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/posts\/22651","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/comments?post=22651"}],"version-history":[{"count":6,"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/posts\/22651\/revisions"}],"predecessor-version":[{"id":22668,"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/posts\/22651\/revisions\/22668"}],"wp:attachment":[{"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/media?parent=22651"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/categories?post=22651"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wjst.de\/blog\/wp-json\/wp\/v2\/tags?post=22651"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}