In a very public, very awkward lesson on the difference between \u201csolving\u201d a problem and simply \u201cfinding\u201d the answer, OpenAI\u2019s latest model, GPT-5, has stirred up controversy.\u00a0<\/p>\n
Top researchers at the company took a premature victory lap on social media, claiming the AI had cracked a set of notoriously difficult mathematical riddles. The celebration, however, was short-lived, earning the AI giant a stern rebuke and a fresh wave of industry ridicule.<\/p>\n
It started October 12, when Sebastien Bubeck, a former Microsoft vice president and now a member of OpenAI\u2019s technical staff, posted on X<\/a>: \u201cGPT-5-pro is superhuman at literature search. It just \u2018solved\u2019 Erd\u0151s problem No. 339 \u2026 by realizing it had been solved 20 years ago.\u201d<\/p>\n The post sounded modest enough, but things soon snowballed. Mathematician Mark Sellke followed up, saying GPT-5 had identified 10 Erd\u0151s problems<\/a> that were \u201clisted as open,\u201d and had made \u201csignificant partial progress\u201d on eleven others.\u00a0<\/p>\n Kevin Weil, OpenAI\u2019s VP of Science, then amplified the excitement with a now-deleted post declaring, \u201cGPT-5 just found solutions to 10 (!) previously unsolved Erd\u0151s problems, and made progress on 11 others. They have all been open for decades.\u201d<\/p>\n It sounded like GPT-5 had made history<\/a>. But the math community wasn\u2019t buying it.<\/p>\n Mathematician Thomas Bloom, who runs the Erd\u0151s Problems website, quickly poured cold water on the celebrations. Responding on X, Bloom wrote<\/a>, \u201cHi, as the owner\/maintainer of http:\/\/erdosproblems.com, this is a dramatic misrepresentation. GPT-5 found references, which solved these problems, that I personally was unaware of.\u201d<\/p>\n He clarified that when a problem is marked as \u201copen\u201d on his website, it simply means he hasn\u2019t seen a paper that solves it, not that the problem is still unsolved. In other words, GPT-5 didn\u2019t actually produce new proofs; it merely discovered papers that already contained the solutions.<\/p>\n Even Google DeepMind\u2019s CEO, Demis Hassabis, chimed in, calling the situation \u201cembarrassing.\u201d<\/a> And Meta\u2019s AI<\/a> chief Yann LeCun took a sharper jab, quipping<\/a>, \u201cHoisted by their own GPTards.\u201d<\/p>\n Despite the drama, the underlying story still holds value. GPT-5 did show impressive capability, not as a mathematician, but as a literature detective. It managed to surface obscure papers that even a domain expert had missed.<\/p>\n As the renowned UCLA mathematician Terence Tao pointed out<\/a>, AI\u2019s real power in math lies not in solving open problems, but in accelerating the grunt work: literature review, pattern search, and hypothesis generation.<\/p>\n That\u2019s exactly what GPT-5 did: it \u201cindustrialized\u201d the research process. Just not in the way its users first claimed.<\/p>\n The incident is another reminder that AI\u2019s achievements should be checked as carefully as a math proof. For OpenAI, it\u2019s a humbling equation: Overconfidence + poor phrasing = one big miscalculation.<\/p>\n GPT-5 didn\u2019t unlock new math, but it did open a discussion about how hype, even when unintentional, can distort scientific credibility.<\/p>\n We live in divisive times. OpenAI has halted the use of <\/strong>Martin Luther King Jr.\u2019s likeness<\/strong><\/a> in Sora after users created fake clips depicting the civil rights leader in \u201cdisrespectful\u201d ways.\u00a0<\/strong><\/p>\n The post Math-etic! GPT-5\u2019s Math \u2018Breakthrough\u2019 Gets F-Minus for Hype<\/a> appeared first on eWEEK<\/a>.<\/p>","protected":false},"excerpt":{"rendered":" In a very public, very awkward lesson on the difference between \u201csolving\u201d a problem and simply \u201cfinding\u201d the answer, OpenAI\u2019s latest model, GPT-5, has stirred up controversy.\u00a0 Top researchers at the company took a premature victory lap on social media, claiming the AI had cracked a set of notoriously difficult mathematical riddles. The celebration, however, […]<\/p>\n","protected":false},"author":0,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[],"class_list":["post-5470","post","type-post","status-publish","format-standard","hentry","category-news"],"_links":{"self":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/posts\/5470"}],"collection":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=5470"}],"version-history":[{"count":0,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/posts\/5470\/revisions"}],"wp:attachment":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5470"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5470"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5470"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}When \u2018open\u2019 doesn\u2019t mean unsolved<\/h2>\n
Subtracting the hype, finding the real story<\/h2>\n
A teachable moment<\/h2>\n