{"id":4225,"date":"2024-09-18T13:00:54","date_gmt":"2024-09-18T13:00:54","guid":{"rendered":"https:\/\/tradetrovex.com\/index.php\/2024\/09\/18\/public-asked-to-help-create-humanitys-last-exam-to-spot-when-ai-achieves-peak-intelligence\/"},"modified":"2024-09-18T13:00:54","modified_gmt":"2024-09-18T13:00:54","slug":"public-asked-to-help-create-humanitys-last-exam-to-spot-when-ai-achieves-peak-intelligence","status":"publish","type":"post","link":"https:\/\/tradetrovex.com\/index.php\/2024\/09\/18\/public-asked-to-help-create-humanitys-last-exam-to-spot-when-ai-achieves-peak-intelligence\/","title":{"rendered":"Public asked to help create \u2018humanity\u2019s last exam\u2019 to spot when AI achieves peak intelligence"},"content":{"rendered":"<p>Scientists are creating \u201chumanity\u2019s last exam\u201d to test AI and see when it has reached expert-level intelligence.<\/p>\n<p>People are being asked to submit their questions and create \u201cthe world\u2019s most difficult <strong>artificial intelligence<\/strong> test\u201d by the Center for AI Safety (CAIS) and Scale AI.<\/p>\n<div class=\"sdc-site-outbrain sdc-site-outbrain--AR_6\">    <\/div>\n<p>\u201cExisting tests now have become too easy and we can no longer track AI developments well, or how far they are from becoming expert-level,\u201d said the quiz creators in a statement about the test.<\/p>\n<p>A few years ago, AI was giving almost random answers to questions on exams \u2013 that\u2019s no longer the case.<\/p>\n<p>Last week, <strong>OpenAI\u2019s<\/strong> newest model, known as OpenAI o1, \u201cdestroyed the most popular reasoning benchmarks\u201d, according to Dan Hendrycks, executive director of CAIS.<\/p>\n<div class=\"ad ad--teads\">        <\/div>\n<p>However, AI still isn\u2019t able to answer difficult research questions and other intellectual questions.<\/p>\n<p>It also appears to score poorly on tests involving planning and visual pattern-recognition puzzles, according to Stanford University\u2019s AI Index Report from April.<\/p>\n<p>Consequently, \u201chumanity\u2019s last exam\u201d will require abstract reasoning to test how clever AI really is.<\/p>\n<p>The submissions shouldn\u2019t be any ordinary quiz questions.<\/p>\n<p>\u201cWe found questions written by undergraduates tend to be too easy for the models,\u201d the creators of the quiz said.<\/p>\n<p>Instead, they recommend that question writers have five or more years of experience in a technical industry job like SpaceX, or are a PhD student or above.<\/p>\n<p>The submissions should be difficult for non-experts to answer and \u201cnot easily answerable via a quick online search\u201d, and trick questions should be avoided.<\/p>\n<p>\u201cAs a rule of thumb, if a randomly selected undergraduate can understand what is being asked, it is likely too easy for the frontier LLMs of today and tomorrow,\u201d said the quiz creators.<\/p>\n<p>People who submit successful questions will be invited as co-authors on the paper and have a chance to win money from a $500,000 (\u00a3378,400) prize pool, with the writers of the best questions earning $5,000 (\u00a33,780) each.<\/p>\n<p>Questions should be submitted by 1 November.<\/p>\n<\/p>\n<div>This post appeared first on sky.com<\/div>\n","protected":false},"excerpt":{"rendered":"<p>Scientists are creating \u201chumanity\u2019s last exam\u201d to test AI and see when it has reached&hellip;<\/p>\n","protected":false},"author":0,"featured_media":4226,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3],"tags":[],"class_list":["post-4225","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tech-news"],"_links":{"self":[{"href":"https:\/\/tradetrovex.com\/index.php\/wp-json\/wp\/v2\/posts\/4225","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/tradetrovex.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/tradetrovex.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/tradetrovex.com\/index.php\/wp-json\/wp\/v2\/comments?post=4225"}],"version-history":[{"count":0,"href":"https:\/\/tradetrovex.com\/index.php\/wp-json\/wp\/v2\/posts\/4225\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/tradetrovex.com\/index.php\/wp-json\/wp\/v2\/media\/4226"}],"wp:attachment":[{"href":"https:\/\/tradetrovex.com\/index.php\/wp-json\/wp\/v2\/media?parent=4225"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/tradetrovex.com\/index.php\/wp-json\/wp\/v2\/categories?post=4225"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/tradetrovex.com\/index.php\/wp-json\/wp\/v2\/tags?post=4225"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}