logo
Google and OpenAI's AI models win milestone gold at global math competition

Google and OpenAI's AI models win milestone gold at global math competition

Reutersa day ago
July 21 (Reuters) - Alphabet's (GOOGL.O), opens new tab Google and OpenAI said their artificial-intelligence models won gold medals at a global mathematics competition, signaling a breakthrough in math capabilities in the race to build powerful systems that can rival human intelligence.
The results marked the first time that AI systems crossed the gold-medal scoring threshold at the International Mathematical Olympiad for high-school students. Both companies' models solved five out of six problems, achieving the result using general-purpose "reasoning" models that processed mathematical concepts using natural language, in contrast to the previous approaches used by AI firms.
The achievement suggests AI is less than a year away from being used by mathematicians to crack unsolved research problems at the frontier of the field, according to Junehyuk Jung, a math professor at Brown University and visiting researcher in Google's DeepMind AI unit.
"I think the moment we can solve hard reasoning problems in natural language will enable the potential for collaboration between AI and mathematicians," Jung told Reuters.
OpenAI's breakthrough was achieved with a new experimental model centered on massively scaling up "test-time compute." This was done by both allowing the model to "think" for longer periods and deploying parallel computing power to run numerous lines of reasoning simultaneously, according to Noam Brown, researcher at OpenAI. Brown declined to say how much in computing power it cost OpenAI, but called it "very expensive."
To OpenAI researchers, it is another clear sign that AI models can command extensive reasoning capabilities that could expand into other areas beyond math.
The optimism is shared by Google researchers, who believe AI models' capabilities can apply to research quandaries in other fields such as physics, said Jung, who won an IMO gold medal as a student in 2003.
Of the 630 students participating in the 66th IMO on the Sunshine Coast in Queensland, Australia, 67 contestants, or about 11%, achieved gold-medal scores.
Google's DeepMind AI unit last year achieved a silver medal score using AI systems specialized for math. This year, Google used a general-purpose model called Gemini Deep Think, a version of which was previously unveiled at its annual developer conference in May.
Unlike previous AI attempts that relied on formal languages and lengthy computation, Google's approach this year operated entirely in natural language and solved the problems within the official 4.5-hour time limit, the company said in a blog post.
OpenAI, which has its own set of reasoning models, similarly built an experimental version for the competition, according to a post by researcher Alexander Wei on social media platform X. He noted that the company does not plan to release anything with this level of math capability for several months.
This year marked the first time the competition coordinated officially with some AI developers, who have for years used prominent math competitions like IMO to test model capabilities. IMO judges certified the results of those companies, including Google, and asked them to publish results on July 28.
"We respected the IMO Board's original request that all AI labs share their results only after the official results had been verified by independent experts and the students had rightly received the acclamation they deserved," Google DeepMind CEO Demis Hassabis said on X on Monday.
OpenAI, which published its results on Saturday and first claimed gold-medal status, said in an interview that it had permission from an IMO board member to do so after the closing ceremony on Saturday.
The competition on Monday allowed cooperating companies to publish results, Gregor Dolinar, president of IMO's board, told Reuters.
Orange background

Try Our AI Features

Explore what Daily8 AI can do for you:

Comments

No comments yet...

Related Articles

OpenAI CEO Sam Altman warns of AI being used for voice fraud in banking
OpenAI CEO Sam Altman warns of AI being used for voice fraud in banking

The Independent

time24 minutes ago

  • The Independent

OpenAI CEO Sam Altman warns of AI being used for voice fraud in banking

Sam Altman, the chief executive of OpenAI, has issued a stark warning to the financial sector, predicting a "significant impending fraud crisis" driven by artificial intelligence. He highlighted AI's growing capability to impersonate human voices, allowing sophisticated bypasses of security protocols to illicitly transfer funds. Speaking at a Federal Reserve conference in Washington, Mr Altman expressed particular alarm over outdated authentication methods. "A thing that terrifies me is apparently there are still some financial institutions that will accept the voiceprint as authentication," he stated. "That is a crazy thing to still be doing. AI has fully defeated that." Voiceprinting, which involves customers uttering a challenge phrase to access accounts, gained popularity over a decade ago, particularly among wealthy bank clients. However, Mr Altman cautioned that AI-generated voice clones, and eventually video clones, are becoming increasingly "indistinguishable from reality," necessitating entirely new verification systems. Michelle Bowman, the Fed's Vice Chair for Supervision and the central bank's top financial regulator, who hosted the discussion with Mr Altman, indicated a willingness to collaborate on solutions. "That might be something we can think about partnering on," she remarked.

Science has proven why your skin wrinkles. Here is what you need to know
Science has proven why your skin wrinkles. Here is what you need to know

The Independent

timean hour ago

  • The Independent

Science has proven why your skin wrinkles. Here is what you need to know

Researchers finally know why our skin wrinkles over time - and Silly Putty can help explain it. Scientists at New York's Binghamton University say experimental evidence shows that it's a similar process to stretching out a favorite hoodie or t-shirt from overuse. Essentially, aging skin stretches in one direction, contracts in another, and then collapses. As you age, the contraction gets bigger, resulting in the formation of the skin folds and creases. 'If you stretch Silly Putty, for instance, it stretches horizontally, but it also shrinks in the other direction — it gets thinner,' Associate Professor of Biomedical Engineering, Guy German, explained in a statement. 'That's what skin does, as well.' Wrinkles start to appear after around the age of 25 years old, according to the Cleveland Clinic. Scientists have long believed that skin wrinkles due to genetics, the effects of disease, and damage from the sun. As you get older, your skin cells are replaced at a slower rate, causing the skin's outer layer to thin and forming wrinkles. Lines in the face, including forehead and frown lines, are largely out of our control, as they're caused by repeated muscle movements. Previous studies, using computational models, have also shown changes in the mechanical properties such as the elasticity and structure of the skin's middle layer during aging. The layer, which contains the proteins elastin and collagen, is the home to hair follicles, blood vessels, and sweat glands. Until now, those changes had never been proven experimentally. 'When I got into this field, that was one of my goals – can I figure out aging?' said German. 'Because if I look at the TV, the radio, online, at shops, I'm being told 1,000 different things about how to improve my skin health, and I want to know what's right and what isn't.' To reach these conclusions, German and his team used a low-force tensometer to stretch out seven tiny strips of skin from people between the ages 16 through 91, simulating the forces the skin naturally experiences. The tensometer tests the maximum force a material can withstand while being pulled or stretched before breaking. The skin was collected through elective surgery or tissue from cadavers. They found that the skin has one set of mechanical properties when you're young. As you age, things get a bit 'wonky,' German noted. 'Things degrade a bit, and it turns out the skin stretches laterally more, which causes the actual wrinkles that form,' said German. 'And the reason why that exists in the first place is that your skin is not in a stress-free state. It's actually stretched a little bit. So there are inherent forces within your skin itself, and those are the driving force towards wrinkles.' The research, which didn't delve into how these forces could be halted, was published recently in the Journal of the Mechanical Behavior of Biomedical Materials. Of course, there are other things that we know affect the skin that can contribute to appearance over time. Spending too much time outside can result in a nasty sun burn, as well as age the skin prematurely with the same effect as aging naturally. 'If you spend your life working outside, you're more likely to have more aged and wrinkled skin than those who are office workers, for example,' German warned.

Wall Street ends mixed; GM slumps as tariffs bite
Wall Street ends mixed; GM slumps as tariffs bite

Reuters

timean hour ago

  • Reuters

Wall Street ends mixed; GM slumps as tariffs bite

July 22 (Reuters) - Wall Street shares ended mixed on Tuesday, with steep losses in General Motors and a gain in Tesla as investors focused on recent and upcoming quarterly reports and watched for signs of progress in U.S. trade discussions. GM (GM.N), opens new tab tumbled after the automaker reported a $1 billion hit from tariffs to its quarterly results, adding more fuel to investor concerns about U.S. President Donald Trump's global trade policy. Shares of Ford Motor (F.N), opens new tab also fell. Tesla (TSLA.O), opens new tab climbed a day before its quarterly report, while Alphabet (GOOGL.O), opens new tab, also reporting on Wednesday, also rose. Optimism about heavy spending on artificial intelligence has underpinned a rally in Wall Street's most valuable companies, with the S&P 500 trading around record highs. "The market is consolidating recent gains and is in a bit of a holding pattern with some huge catalysts over the next week or two, including the August 1 tariff deadline and a lot of important Magnificent Seven earnings," said Ross Mayfield, an investment strategy analyst at Baird. Other Big Tech stocks lost ground, with Meta Platforms (META.O), opens new tab and Microsoft (MSFT.O), opens new tab both closing lower. Shares of RTX (RTX.N), opens new tab dropped after the aerospace and defense giant t from Trump's trade war despite strong demand for its engines and aftermarket services. Lockheed Martin (LMT.N), opens new tab tumbled after its quarterly profit plunged by about 80%. U.S. trade policy remains a major point of uncertainty for investors and companies as Trump's self-imposed August 1 deadline for many countries to reach agreements with the White House approaches. U.S. Treasury Secretary Scott Bessent said he would meet his Chinese counterpart next week to discuss an extension to the August 12 deadline set for tariffs on imports from China. Other trade negotiations appeared stalled, with optimism for a breakthrough deal with India waning and EU officials weighing countermeasures against the United States. According to preliminary data, the S&P 500 (.SPX), opens new tab gained 4.30 points, or 0.07%, to end at 6,309.90 points, while the Nasdaq Composite (.IXIC), opens new tab lost 81.24 points, or 0.39%, to 20,892.93. The Dow Jones Industrial Average (.DJI), opens new tab rose 175.77 points, or 0.40%, to 44,498.84. Philip Morris (PM.N), opens new tab slumped after reporting second-quarter revenue below expectations, as shipments of its ZYN nicotine pouches disappointed investors. Analysts on average expected S&P 500 companies to report a 7% increase in earnings for the second quarter, with technology heavyweights driving much of that gain, according to LSEG I/B/E/S. After last week's mixed economic data, traders have all but ruled out an interest-rate cut from the U.S. Federal Reserve at next week's policy meeting. They now see about a 60% chance of a reduction in September, according to the CME's FedWatch tool.

DOWNLOAD THE APP

Get Started Now: Download the App

Ready to dive into a world of global content with local flavor? Download Daily8 app today from your preferred app store and start exploring.
app-storeplay-store