Key Takeaways OpenAI executives falsely claimed GPT-5 solved previously unsolved mathematical problems, sparking industry backlash. The AI model had only found existing research papers that mathematic...
OpenAI faced sharp criticism from industry leaders and mathematicians this week after executives prematurely claimed the company's GPT-5 model had solved longstanding mathematical problems, only to retract the statements when experts revealed the AI had merely located existing research papers.
The controversy erupted on October 18 when Kevin Weil, OpenAI's Vice President, posted on X that GPT-5 had found solutions to 10 previously unsolved Erdős problems and made progress on 11 others.
Erdős problems are famous mathematical conjectures posed by Hungarian mathematician Paul Erdős, many of which have remained unsolved for decades.
The celebration was short-lived. Thomas Bloom, who maintains the Erdos Problems website, quickly corrected the record, calling Weil's post "a dramatic misrepresentation" of what actually occurred.
"GPT-5 found references, which solved these problems that I personally was unaware of," Bloom wrote on X. "The 'open' status only means I personally am unaware of a paper that solves it."
In other words, GPT-5 had not independently solved any mathematical problems. Instead, it had performed sophisticated literature searches to find academic papers containing solutions that Bloom hadn't cataloged on his website.
The misstep drew immediate ridicule from OpenAI's competitors. Google DeepMind CEO Demis Hassabis responded bluntly to the claims, posting "this is embarrassing" on X.
Meta's Chief AI Scientist Yann LeCun was more pointed in his criticism, quipping that OpenAI had been "Hoisted by their own GPTards" — a play on the phrase "hoisted by your own petard," suggesting the company had become a victim of its own hype.
Following the backlash, OpenAI researcher Sebastien Bubeck, who had also promoted the supposed breakthrough, deleted his original post and issued an apology.
"I deleted the post, I didn't mean to mislead anyone obviously, I thought the phrasing was clear, sorry about that," Bubeck wrote on X. "Only solutions in the literature were found that's it, and I find this very accelerating because I know how hard it is to search the literature."
Weil's original post was also deleted, and other OpenAI researchers who had echoed the claims retracted or amended their statements.
The incident highlights mounting pressure within the artificial intelligence sector to demonstrate groundbreaking capabilities.
With billions of dollars at stake and intense competition between companies like OpenAI, Google, and Meta, the rush to announce achievements has raised concerns about scientific rigor and verification.
Industry observers noted the irony of the situation: OpenAI had achieved a genuine mathematical milestone in July when an experimental model earned a gold-medal score at the International Mathematical Olympiad. The subsequent unforced error regarding Erdős problems appeared particularly puzzling, given the earlier success.
Despite the embarrassment, the episode did reveal GPT-5's practical utility as a research tool.
Renowned mathematician Terence Tao has noted that AI's most immediate potential in mathematics may lie not in solving the toughest open problems, but in speeding up tedious tasks like literature searches, exactly what GPT-5 actually accomplished.
Key takeaways OpenAI has signed a multi-year, $10 billion agreement with AI chipmaker Cerebras Systems to secure computing infrastructure. The deal will deliver 750 megawatts of computing power throug...
KEY TAKEAWAYS: xAI implemented restrictions preventing Grok from editing images of real people in revealing clothing after global backlash California Attorney General Rob Bonta launched investigation...
Key takeaways TSMC reported fourth-quarter net income of NT$505.74 billion, beating analyst expectations and setting a new record. Fourth-quarter revenue reached NT$1.046 trillion, up 20.5% year over...
