Humans beat AI gold-level score at top maths contest
Neither model 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 maths 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 US 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 percent of human contestants won gold-level medals, and five received perfect scores of 42 points.
US ChatGPT maker OpenAI said that its experimental reasoning model had 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 wrote on social media.
'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 British city of Bath, 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.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Express Tribune
10 hours ago
- Express Tribune
Opal AI test-launched by Google: Here's how it can generate apps without coding, with a simple prompt
Google is piloting a new AI-powered coding tool called Opal, designed to help users build mini web applications by describing what they want to create in plain language. The tool, now accessible in the United States via Google Labs, is the latest in a growing wave of so-called "vibe-coding" platforms targeting both developers and non-technical users, as reported by TechCrunch. Opal enables users to either start from scratch using text-based instructions or customise existing apps from a shared gallery. Once a prompt is submitted, Google's internal models generate the application's structure. Users are then presented with a visual workflow panel that outlines each step in the process - including input, output, and generation phases. 🚨 NEW LABS EXPERIMENT 🚨 Introducing Opal, our new way to help you build and share AI mini-apps by linking together prompts, models, and tools— all while using simple, natural language (without a single line of code 🤯) Now available in US-only public beta! Learn more ⬇️… — Google Labs (@GoogleLabs) July 24, 2025 Each stage can be clicked to reveal and edit the underlying prompt, while additional steps can be added manually through Opal's toolbar. Finished applications can be published online and shared via a direct link, with others able to test the app using their own Google accounts. While Google already offers an AI Studio for developers to build apps through prompt engineering, Opal appears to mark a broader push towards accessibility and design-driven prototyping. Its visual-first interface and ease of use suggest the company is aiming to appeal to hobbyists, creatives, and aspiring app builders without coding experience. The move brings Google into direct competition with companies such as Canva, Figma, and Replit, all of which have introduced tools that lower the barrier to entry for software creation. Startups in the space, including Lovable and Cursor, have recently drawn significant investor and acquisition interest amid rising demand for generative AI solutions. Opal is currently in an experimental phase and forms part of Google's wider efforts to test emerging AI technologies through its Labs division.
Express Tribune
11 hours ago
- Express Tribune
Microsoft moves towards AI transformation following 9000 job cuts in 2025
Microsoft CEO Satya Nadella has confirmed the company's commitment to an AI transformation following the layoff of 9000 employees in July 2025. These cuts are part of a broader restructuring, with over 15,000 roles removed this year. Microsoft has stated it is not facing financial issues but is shifting priorities, with significant investment in AI infrastructure, including data centres and Copilot tools. In May, Microsoft laid off 6000 engineers and developers, while the July reductions focused on sales and support teams. The company is using third-party representatives and AI tools that currently write 30 per cent of its code, reducing the need for certain roles. Addressing the layoffs in a company-wide memo shared by Business Insider, Nadella described the decisions as 'among the most difficult'. He added that the company's overall headcount is 'relatively unchanged', despite the nearly four per cent workforce reduction. Nadella acknowledged the uncertainty created by these changes, stating, 'It might feel messy at times, but transformation always is.' He encouraged employees to view the shift positively, highlighting benefits such as 'more speed, more scale, more impact' and describing a future where AI researchers and coding assistants are readily accessible to employees. Microsoft has identified security, quality, and AI as its current priorities, with AI described as a key factor in the company's future operations. The move aligns with broader industry trends, with companies including Google, IBM, and Netmarble also adopting AI tools to restructure operations and reduce costs.
Express Tribune
a day ago
- Express Tribune
Human mathematicians best AI in competition
Humans beat generative AI models made by Google and OpenAI at a top international mathematics competition, despite the programs reaching gold-level scores for the first time, reported AFP. Neither model 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 that an advanced version of its Gemini chatbot had solved five out of the six maths 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 US 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 per cent of human contestants won gold-level medals, and five received perfect scores of 42 points. US ChatGPT maker OpenAI said that its experimental reasoning model had 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 wrote on social media. "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 British city of Bath, 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.
