{"id":12054,"date":"2024-03-17T14:50:12","date_gmt":"2024-03-17T14:50:12","guid":{"rendered":"https:\/\/edusign.com\/?p=12054"},"modified":"2024-02-22T09:47:37","modified_gmt":"2024-02-22T09:47:37","slug":"openai-has-launched-a-second-tool-to-complement-chatgpt-and-help-teachers-detect-cheating","status":"publish","type":"post","link":"https:\/\/edusign.com\/blog\/openai-has-launched-a-second-tool-to-complement-chatgpt-and-help-teachers-detect-cheating\/","title":{"rendered":"OpenAI has launched a second tool to complement ChatGPT – and help teachers detect cheating."},"content":{"rendered":"\n

Since the arrival <\/strong>of ChatGPT, an artificial intelligence-generated <\/strong>text tool, opinions have been mixed<\/strong>, particularly in the field of education<\/a><\/strong>. Capable of generating <\/strong>a variety of textual content<\/strong>, this tool <\/strong>can also be used for malicious <\/strong>purposes such as disinformation <\/strong>or cheating<\/a>.<\/p>\n\n\n\n

OpenAI, the company <\/strong>behind the famous ChatGPT, has launched <\/strong>a tool capable of detecting whether a text has been written <\/strong>by an artificial intelligence <\/strong>(AI) or a human. This detector <\/strong>is capable of predicting <\/strong>with what probability <\/strong>a text, of at least 1,000 characters<\/strong>, has been imagined by AI rather <\/strong>than by a human being<\/strong>.<\/p>\n\n\n\n

How does it work?<\/h2>\n\n\n\n

Once a text of at least 1000 characters <\/strong>has been entered into the tool, the software <\/strong>will label the document according <\/strong>to the probability that the AI has intervened<\/strong>. There are five levels of probability<\/strong>: very unlikely <\/strong>(10% chance), unlikely <\/strong>(between 10 and 45%), unclear <\/strong>(between 45 and 90%), maybe <\/strong>(between 90 and 98%), or probably <\/strong>(over 98%). OpenAI nevertheless <\/strong>estimates that 9% of texts are incorrectly considered <\/strong>to have been written by AI.<\/p>\n\n\n\n

But the company <\/strong>has warned users <\/strong>on its blog that the tool is not entirely <\/strong>reliable, and the creators want feedback <\/strong>on how their new tool works.<\/p>\n\n\n\n

We’ve tested the tool for you:<\/h2>\n\n\n\n

Some students <\/a><\/strong>may be tempted <\/strong>to use ChatGPT to write assignments<\/strong>.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

To test the tool’s fallibility <\/strong>and see its limits, we generated <\/strong>an essay on ChatGPT and then used <\/strong>a tool to paraphrase the text in order to blur the lines <\/strong>of a possible text generated <\/strong>by an artificial intelligence<\/strong>.<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

After paraphrasing <\/strong>the text, AI Classifier Text still believes <\/strong>that the text is generated <\/strong>by an artificial intelligence. In conclusion<\/strong>, the new tool achieves a positive breakthrough <\/strong>in preventing cheating<\/a><\/strong>. A student <\/a>who still wants <\/strong>to use ChatGPT by paraphrasing <\/strong>his sentences with a tool will still have <\/strong>to work manually and therefore <\/strong>use his knowledge<\/strong>.<\/p>\n\n\n\n

How to avoid cheating without tools?<\/h2>\n\n\n\n

There are several <\/strong>ways to help teachers detect <\/strong>cheating<\/a>:<\/p>\n\n\n\n

    \n
  1. Monitor suspicious <\/strong>behavior: Teachers can be alerted by behaviors <\/strong>such as duplication of answers, lack of engagement during the exam, frequent <\/strong>consultation of notes<\/strong>, etc.<\/li>\n\n\n\n
  2. Use proctored online <\/a>exams<\/strong>: Online <\/a>exams can be monitored <\/strong>for suspicious behavior<\/strong>, such as copying text or consulting <\/strong>websites.<\/li>\n\n\n\n
  3. Diversify exam questions: Teachers can diversify <\/strong>exam questions to prevent <\/strong>students <\/a>from sharing answers.<\/li>\n\n\n\n
  4. Assign random <\/strong>topics: Teachers can assign <\/strong>random topics to prevent students <\/a><\/strong>from being unable to prepare in advance <\/strong>by sharing answers<\/strong>.<\/li>\n<\/ol>\n","protected":false},"excerpt":{"rendered":"

    Since the arrival of ChatGPT, an artificial intelligence-generated text tool, opinions have been mixed, particularly in the field of education. Capable of generating a variety of textual content, this tool can also be used for malicious purposes such as disinformation or cheating. OpenAI, the company behind the famous ChatGPT, has launched a tool capable of […]<\/p>\n","protected":false},"author":15,"featured_media":12055,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"content-type":""},"categories":[79],"tags":[],"acf":[],"_links":{"self":[{"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/posts\/12054"}],"collection":[{"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/comments?post=12054"}],"version-history":[{"count":2,"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/posts\/12054\/revisions"}],"predecessor-version":[{"id":12148,"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/posts\/12054\/revisions\/12148"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/media\/12055"}],"wp:attachment":[{"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/media?parent=12054"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/categories?post=12054"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/edusign.com\/wp-json\/wp\/v2\/tags?post=12054"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}