Google and OpenAI's AI models win milestone gold at global math competition
(Reuters) -Alphabet's 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 mathematiciansto 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.
The same idea 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.
"Werespected 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.
However, OpenAI, which did not work with the IMO, self-published its results on Saturday, allowing it to be first among AI firms to claim gold-medal status.
In turn, the competition on Monday allowed cooperating companies to publish results, Gregor Dolinar, president of IMO's board, told Reuters.
(Reporting by Kenrick Cai in San Francisco and Jaspreet Singh in Bengaluru; Editing by Matthew Lewis)
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
The Star
42 minutes ago
- The Star
Alphabet hit with EU antitrust complaint by six digital rights groups
FILE PHOTO: The new Google logo is seen in this illustration taken May 13, 2025. REUTERS/Dado Ruvic/Illustration/File Photo BRUSSELS (Reuters) -Alphabet was targeted with an EU antitrust complaint from six human and digital rights groups on Thursday which urged EU regulators to investigate whether the tech giant complies with legislation requiring it to make it easier for users to uninstall software apps. The EU's Digital Markets Act (DMA), which took effect two years ago, sets out a list of dos and don'ts for seven Big Tech companies including Alphabet unit Google, in an attempt to curb their power and give rivals more room and users more choice. British human rights organisation ARTICLE 19, European Digital Rights (EDRi), Free Software Foundation Europe (FSFE), Gesellschaft fur Freiheitsrechte (GFF), Homo Digitalis, and said Alphabet allegedly has not complied with the DMA. They said the alleged violation concerned a DMA requirement that gatekeepers or companies which provide a core platform service to business users, shall technically enable users to easily uninstall software applications on the gatekeeper's operating system. "Alphabet has designed its Core Platform Service Android in a way to hide from end users the possibility to disable its own pre-installed gatekeeper apps," they said in their complaint. "What is more, Alphabet goes to great length to scare away end users who have found that possibility against all odds of actually disabling Google's pre-installed apps," they said. The group called on the European Commission to investigate the issue. Alphabet refuted the allegations. "It is easy to uninstall apps on Android devices, so this complaint does not represent a genuine user concern. Other regulators, including the CMA, have previously dismissed this complaint," a Google spokesperson said. The CMA is the British competition watchdog. The Commission confirmed receipt of the complaint, saying it is currently assessing it under its standard procedures. (Reporting by Foo Yun Chee; Editing by Susan Fenton)
New Straits Times
2 hours ago
- New Straits Times
Dotcom lessons loom over AI-fuelled Wall Street surge
WALL Street's concentration in the red-hot tech sector is, by some measures, greater than it has ever been, eclipsing levels hit during the 1990s dotcom bubble. But does this mean history is bound to repeat itself? The growing concentration in United States equities instantly brings to mind the Internet and communications frenzy of the late 1990s. The tech-heavy Nasdaq peaked in March 2000 before cratering 65 per cent over the following 12 months. And it didn't revisit its previous high for 14 years. It seems unlikely that we will see a repeat of this today, right? Maybe. The market's reaction function appears to be different from what it was during the dotcom boom and bust. Just look at the current rebound from its post-"Liberation Day" tariff slump in early April — one of the fastest on record — or its rally during the Covid-19 pandemic. But despite all of these differences, there are also some worrying parallels. Investors would do well to keep both in mind. The most obvious similarity between these two periods is the concentration of tech and related industries in US equity markets. The broad tech sector now accounts for 34 per cent of the S&P 500's market cap, according to some data, exceeding the previous record of 33 per cent set in March 2000. Of the top 10 companies by market capitalisation today, eight are tech or communications behemoths. They include Apple, Amazon, Alphabet, Meta, Microsoft, Nvidia and Tesla — as well as Berkshire Hathaway and JPMorgan. By contrast, only five of the 10 biggest companies in 1999 were tech firms. The other five were General Electric, Citi, Exxon, Walmart and Home Depot. Plus, the top 10 companies' footprint in the S&P 500 today is much larger than it was back then. The combined market cap of the top 10 today is almost US$22 trillion, or 40 per cent of the index's total, significantly higher than the comparable 25 per cent in 1999. This all reflects the fact that technology plays a much bigger role in the US economy today. By some measures, the current tech boom, driven in part by enthusiasm for artificial intelligence (AI), is more extreme than the information technology bubble of the late 1990s. As Torsten Slok, chief economist at Apollo Global Management, points out, the 12-month forward earnings valuation of today's top 10 stocks in the S&P 500 is higher than it was 25 years ago. However, it's worth remembering that the dotcom bubble was characterised by a frenzy of public offerings and a raft of companies with shares valued at triple-digit multiples of future earnings. That's not the case today. While the S&P tech sector is trading at 29.5 times forward earnings today, which is high by historical standards, this is nowhere near the peak of almost 50 times recorded in 2000. Similarly, the S&P 500 and Nasdaq are currently trading around 22 and 28.5 times forward earnings, compared with the dotcom peaks of 24.5 and over 70 times, respectively. With all that being said, a meaningful, prolonged market correction cannot be ruled out, especially if AI-driven growth isn't delivered as quickly as investors expect. AI, the new driver of technological development, will require vast capital outlays, especially on data centres, which may mean that earnings and share price growth in tech could slow in the short run. According to Morgan Stanley, the transformative potential of generative AI will require roughly US$2.9 trillion of global data centre spending through 2028, comprising US$1.6 trillion on hardware like chips and servers and US$1.3 trillion on infrastructure. That means investment needs of over US$900 billion in 2028, they reckon. For context, combined capital expenditure by all S&P 500 companies last year was around US$950 billion. Wall Street analysts are well aware of these figures, which suggests that at least some percentage of these huge sums should be factored into current share prices and expected earnings, but what if the benefits of AI take longer to deliver? Or what if an upstart (remember China's DeepSeek?) dramatically shifts growth expectations for a major component of the index, like US$4-trillion chipmaker Nvidia? Of course, technology is so fundamental to today's society and economy that it's difficult to imagine its market footprint shrinking too much, for too long, as this raises the inevitable question of where investor capital would go.
The Star
4 hours ago
- The Star
Wall St set for mixed start as investors digest Alphabet, Tesla results
A woman walks down Wall Street in New York City, U.S., April 8, 2025. REUTERS/Kylie Cooper/File Photo WALL Street was on track for a mixed open on Thursday as investors digested uneven earnings from megacaps like Alphabet and Tesla and monitored progress in U.S. trade negotiations. Alphabet rose 3.4% in premarket after the Google parent raised its 2025 capital spending forecast by $10 billion to $85 billion, shrugging off trade jitters, while electric vehicle maker Tesla tumbled 6.1% as CEO Elon Musk warned of "a few rough quarters" due to cuts in EV incentives. At 8:43 a.m. ET, S&P 500 E-minis were up 0.5 points, or 0.01%, Nasdaq 100 E-minis were up 42 points, or 0.18%, and Dow E-minis were down 292 points, or 0.65%. UnitedHealth fell 4.5%. The insurer said it is cooperating with the Department of Justice's formal criminal and civil requests following reports of investigations into its Medicare participation. The S&P 500 and the tech-heavy Nasdaq soared to record closes on Wednesday as investors cheered reports of an imminent trade deal between Washington and the European Union. Meanwhile, the Dow closed over 1.1% higher, just below its all-time peak. An EU spokesperson on Thursday signaled that a deal was "within reach", which, as per diplomats, would result in broad 15% import tariffs on the 27-member bloc. Anticipation of further trade pacts was also reinforced by President Donald Trump's announcement of a deal with Japan on Tuesday, cutting import levies on the Asian country to just 15%. Meanwhile, China and South Korea are racing to strike agreements to dodge Trump's hefty duties. "A U.S. tariff agreement with Japan has increased market confidence that the worst of the global trade conflict could be over, adding to hopes of a deal with the European Union," said Mark Haefele, chief investment officer, UBS Global Wealth Management. Among other stocks, American Airlines fell 3.3% after forecasting a bigger-than-expected third-quarter loss, hurt by sluggish domestic travel demand. IBM slid 6.3% as its second-quarter earnings failed to impress investors, especially due to its lower-than-expected sales in its mainstay software segment. Honeywell slipped 2.8% despite raising its annual forecasts after beating Wall Street expectations for second-quarter results. Shares of ServiceNow jumped 7.4% after the software firm raised its annual subscription revenue forecast. Markets were also monitoring developments after the White House surprised investors that Trump - fresh from stepping up his criticism of Federal Reserve Chair Jerome Powell - would pay a visit to the U.S. central bank's headquarters later in the day. With the Fed widely expected to keep rates steady at next week's meeting, traders are now eyeing a 62% chance of a September rate cut, according to CME's FedWatch tool. A Labor Department's report showed jobless claims for the week ended July 18 stood at 217,000 versus an estimate of 235,000, signalling resilience in the job market. S&P Global's flash PMI data will be released at 9:45 a.m. ET. - Reuters
