logo
Humans beat AI at annual math Olympiad, but the machines are catching up

Humans beat AI at annual math Olympiad, but the machines are catching up

Yahoo5 days ago
Sydney — Humans beat generative AI models made by Google and OpenAI at a top international mathematics competition, but the programs reached gold-level scores for the first time, and the rate at which they are improving may be cause for some human introspection. Neither of the AI models scored full marks — unlike five young people at the International Mathematical Olympiad (IMO), a prestigious annual competition where participants must be under 20 years old. Google said Monday that an advanced version of its Gemini chatbot had solved five out of the six math problems set at the IMO, held in Australia's Queensland this month. "We can confirm that Google DeepMind has reached the much-desired milestone, earning 35 out of a possible 42 points - a gold medal score," the U.S. tech giant cited IMO president Gregor Dolinar as saying. "Their solutions were astonishing in many respects. IMO graders found them to be clear, precise and most of them easy to follow." Around 10% of human contestants won gold-level medals, and five received perfect scores of 42 points. U.S. ChatGPT maker OpenAI said its experimental reasoning model had also scored a gold-level 35 points on the test. The result "achieved a longstanding grand challenge in AI" at "the world's most prestigious math competition," OpenAI researcher Alexander Wei said in a social media post. "We evaluated our models on the 2025 IMO problems under the same rules as human contestants," he said. "For each problem, three former IMO medalists independently graded the model's submitted proof." Google achieved a silver-medal score at last year's IMO in the city of Bath, in southwest England, solving four of the six problems. That took two to three days of computation — far longer than this year, when its Gemini model solved the problems within the 4.5-hour time limit, it said. The IMO said tech companies had "privately tested closed-source AI models on this year's problems," the same ones faced by 641 competing students from 112 countries. "It is very exciting to see progress in the mathematical capabilities of AI models," said IMO president Dolinar. Contest organizers could not verify how much computing power had been used by the AI models or whether there had been human involvement, he noted.
In an interview with CBS' 60 Minutes earlier this year, one of Google's leading AI researchers predicted that within just five to 10 years, computers would be made that have human-level cognitive abilities — a landmark known as "artificial general intelligence."
Google DeepMind CEO Demis Hassabis predicted that AI technology was on track to understand the world in nuanced ways, and to not only solve important problems, but even to develop a sense of imagination, within a decade, thanks to an increase in investment.
"It's moving incredibly fast," Hassabis said. "I think we are on some kind of exponential curve of improvement. Of course, the success of the field in the last few years has attracted even more attention, more resources, more talent. So that's adding to the, to this exponential progress."
Detroit lawnmower gang still going strong after 15 years
Legendary singer Ozzy Osbourne dies at 76
Sneak peek: The Case of the Black Swan (Part 1)
Solve the daily Crossword
Orange background

Try Our AI Features

Explore what Daily8 AI can do for you:

Comments

No comments yet...

Related Articles

Android phones helping detect potential earthquakes
Android phones helping detect potential earthquakes

Yahoo

timean hour ago

  • Yahoo

Android phones helping detect potential earthquakes

(NewsNation) — More and more every day, it seems like smartphones can expand our knowledge on everything, including detecting potential earthquakes. Earthquakes? Yes. Recently, researchers from Google and partner institutions shared results from the Android Earthquake Alerts system. The AEA, over the last three years, has expanded earthquake warning coverage from 250 million people to 2.5 billion in 98 countries. The system sends a signal to Google's earthquake detection server, and the possible location where the shaking occurs. It then analyzes data from phones to confirm that an earthquake is happening, while also estimating its location and magnitude. Two alerts are then sent out: BeAware and TakeAction. Alaska is the most earthquake-prone state. Here is why Wednesday's earthquake was notable 'The system has now detected over 18,000 earthquakes, from small tremors of M1.9 to major quakes reaching M7.8,' according to the research. 'For the events significant enough to warn people, alerts were issued for over 2000 earthquakes, culminating in 790 million alerts being sent to phones worldwide.' 'The impact has been a ~10x change in the number of people with access to EEW systems.' Dating back to March 31, 2024, AEA has issued alerts to Android phones for a total of 1279 events that were detected. Only three were false alarms, with two resulting from thunderstorms. Android phones make up more than 70% of the world's smartphones as of July 2025. Copyright 2025 Nexstar Media, Inc. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed. Solve the daily Crossword

Google failed to warn 10 million of Turkey earthquake
Google failed to warn 10 million of Turkey earthquake

Yahoo

time2 hours ago

  • Yahoo

Google failed to warn 10 million of Turkey earthquake

Google has admitted its earthquake early warning system failed to accurately alert people during Turkey's deadly quake of 2023. Ten million people within 98 miles of the epicentre could have been sent Google's highest level alert - giving up to 35 seconds of warning to find safety. Instead, only 469 "Take Action" warnings were sent out for the first 7.8 magnitude quake. Google told the BBC half a million people were sent a lower level warning, which is designed for "light shaking", and does not alert users in the same prominent way. The tech giant previously told the BBC the system had "performed well". The system works on Android devices, which make up more than 70% of the phones in Turkey. More than 55,000 people died when two major earthquakes hit south-east Turkey on 6 February 2023, more than 100,000 were injured. Many were asleep in buildings that collapsed around them when the tremors hit. Google's early warning system was in place and live on the day of the quakes – however it underestimated how strong the earthquakes were. "We continue to improve the system based on what we learn in each earthquake", a Google spokesperson said. How it works Google's system, named Android Earthquake Alerts (AEA), is able to detect shaking from a vast number of mobile phones that use the Android operating system. Because earthquakes move relatively slowly through the earth, a warning can then be sent out. Google's most serious warning is called "Take Action", which sets off a loud alarm on a user's phone - overriding a Do Not Disturb setting - and covering their screen. This is the warning that is supposed to be sent to people when stronger shaking is detected that could threaten human life. AEA also has a less serious "Be Aware" warning, designed to inform users of potential lighter shaking - a warning that does not override a device on Do Not Disturb. The Take Action alert was especially important in Turkey due to the catastrophic shaking and because the first earthquake struck at 04:17, when many users would have been asleep. Only the more serious alert would have woken them. In the months after the earthquake the BBC wanted to speak to users who had been given this warning - initially with aims to showcase the effectiveness of the technology. But despite speaking to people in towns and cities across the zone impacted by the earthquake, over a period of months, we couldn't find anyone who had received a more serious Take Action notification before the quake struck. We published our findings later that year. 'Limitations' Google researchers have written in the Science journal details of what went wrong, citing "limitations to the detection algorithms". For the first earthquake, the system estimated the shaking at between 4.5 and 4.9 on the moment magnitude scale (MMS) when it was actually a 7.8. A second large earthquake later that day was also underestimated, with the system this time sending Take Action alerts to 8,158 phones and Be Aware alerts to just under four million users. After the earthquake Google's researchers changed the algorithm, and simulated the first earthquake again. This time, the system generated 10 million Take Action alerts to those at most risk – and a further 67 million Be Aware alerts to those living further away from the epicentre "Every earthquake early warning system grapples with the same challenge - tuning algorithms for large magnitude events," Google told the BBC. But Elizabeth Reddy, assistant professor at Colorado School of Mines, says it is concerning it took more than two years to get this information. "I'm really frustrated that it took so long," she said "We're not talking about a little event - people died - and we didn't see a performance of this warning in the way we would like." Google says the system is supposed to be supplementary and is not a replacement for national systems. However some scientists worry countries are placing too much faith in tech that has not been fully tested. "I think being very transparent about how well it works is absolutely critical," Harold Tobin, director of the Pacific Northwest Seismic Network, told the BBC. "Would some places make the calculation that Google's doing it, so we don't have to?" Google researchers say post-event analysis has better improved the system - and AEA has pushed out alerts in 98 countries. The BBC has asked Google how AEA performed during the 2025 earthquake in Myanmar, but has yet to receive a response. How a grieving mother exposed the truth of Turkey's deadly earthquake Beverley man remembers family lost in Turkey quake Sign up for our Tech Decoded newsletter to follow the world's top tech stories and trends. Outside the UK? Sign up here.

Samsung reveals a mysterious $16.5 billion chip deal.
Samsung reveals a mysterious $16.5 billion chip deal.

The Verge

time2 hours ago

  • The Verge

Samsung reveals a mysterious $16.5 billion chip deal.

Chip race: Microsoft, Meta, Google, and Nvidia battle it out for AI chip supremacy See all Stories Posted Jul 28, 2025 at 3:04 AM UTC Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates. Richard Lawler Posts from this author will be added to your daily email digest and your homepage feed. See All by Richard Lawler Posts from this topic will be added to your daily email digest and your homepage feed. See All Business Posts from this topic will be added to your daily email digest and your homepage feed. See All News Posts from this topic will be added to your daily email digest and your homepage feed. See All Samsung Posts from this topic will be added to your daily email digest and your homepage feed. See All Tech

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