Humans beat AI gold-level score at top maths contest
Empower your mind, elevate your skills
Humans beat generative AI models made by Google and OpenAI at a top international mathematics competition, despite the programmes reaching gold-level scores for the first time.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."It is very exciting to see progress in the mathematical capabilities of AI models," said IMO president Dolinar.Contest organisers could not verify how much computing power had been used by the AI models or whether there had been human involvement, he cautioned.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Hindustan Times
4 hours ago
- Hindustan Times
ChatGPT therapy chats are not private, warns OpenAI CEO Sam Altman
More people are using ChatGPT like a therapist, but that doesn't mean it's private. OpenAI CEO Sam Altman says those kinds of chats don't have the same legal protections you get with real therapists, doctors, or lawyers. OpenAI says deleted chats from Free, Plus, and Pro users are wiped within 30 days unless they're legally required to keep them for "legal or security reasons."(AP) Altman told podcaster Theo Von, 'So if you go talk to ChatGPT about your most sensitive stuff and then there's like a lawsuit or whatever, we could be required to produce that, and I think that's very screwed up.' He went on, 'Right now, if you talk to a therapist or a lawyer or a doctor about those problems, there's like legal privilege for it — there's doctor-patient confidentiality, there's legal confidentiality,' according to Business Insider report. 'We haven't figured that out yet for when you talk to ChatGPT.' Altman said there should be the 'same concept of privacy for your conversations with AI that we do with a therapist' and that it should be 'addressed with some urgency.' Also Read: 'Will AI replace lawyers? Law intern asks to use ChatGPT for witness analysis, gets hard copies instead Youngsters turning to ChatGPT for therapy He said a growing number of people — especially younger users — are turning to ChatGPT for therapy, life advice, or help with relationships. Altman said, 'No one had to think about that even a year ago, and now I think it's this huge issue of like, 'How are we gonna treat the laws around this?'' Unlike end-to-end encrypted apps like WhatsApp or Signal, OpenAI can read your conversations with ChatGPT. Employees sometimes look at chats to improve the AI or to watch for misuse. OpenAI says deleted chats from Free, Plus, and Pro users are wiped within 30 days unless they're legally required to keep them for "legal or security reasons." Back in June, The New York Times and other media outlets asked a court to force OpenAI to save all user chats, even deleted ones, as part of a copyright lawsuit. OpenAI is now appealing that court order.
Time of India
6 hours ago
- Time of India
Shop or drop: What is India Inc.'s take on AI agents?
AI agents are no longer just assistants—they're turning into autonomous actors. With the global launch of OpenAI's ChatGPT Agents and Perplexity's Comet in July 2025, Indian enterprises are now piloting systems that can execute full workflows—research, report generation, and task automation—without manual triggers. The technology has arrived. The promise is real. But the question remains: Is India Inc. ready to trust it? ChatGPT Agents vs. Perplexity Comet: What's the Difference? While both platforms aim to empower autonomous AI execution, their design philosophies differ. ChatGPT Agents are goal-driven workers that operate across tools and APIs. Enterprises can define custom instructions, connect to data sources, and delegate end-to-end tasks—from summarising legal documents to orchestrating backend operations. The emphasis is on flexible, programmable workflows, often embedded within corporate tools like Slack or Comet, by contrast, leans heavily into research and reasoning. It's designed as an autonomous knowledge assistant that continuously searches, validates, and synthesises information from the web to generate insights or reports—with citation and traceability baked in. Enterprise pilots begin, but trust lags behind. According to a new PwC survey, 79% of global executives are already piloting AI agents, and two-thirds report measurable gains in efficiency. Yet only 36% say they're confident in managing the risks. In India, early adopters like ABB Energy Industries and Raychem RPG are already experimenting—but with guardrails. 'Absolutely, my team and I have actively explored ChatGPT's AI agents and similar autonomous AI solutions across pilot projects and internal innovation sandboxes. My first impression centred on both their versatility and the speed at which they could deliver actionable results from complex datasets. The agentic model, especially when layered atop a robust, single-source-of-truth data platform, revealed remarkable potential to automate routine decisions and even initiate complex analytic tasks end-to-end with minimal human input,' said Chandan Vijay, Chief Data Officer at ABB Energy Industries. 'While the technology is powerful, its real impact emerges only when it has access to well-governed, high-quality data—underscoring the immense value of our investment in foundational data architecture,' adds Vijay. At Raychem RPG, Chief Digital & Information Officer, Mehjabeen Taj Aalam struck a similar chord: 'Yes, we've started experimenting with agentic AI tools, including those from OpenAI and other platforms. Our first impression? Equal parts fascination and caution. The ability of these agents to not just respond but initiate actions across systems is a game-changer. But it also forces you to rethink control, context, and trust in a very fundamental way.' CIOs prioritise internal ops amid regulatory fog Most Indian enterprises are deploying agents in internal operations and analytics—report automation, anomaly detection, and consent workflows—where regulatory and reputational risks are lower. 'While the technology is powerful, its real impact emerges only when it has access to well-governed, high-quality data—underscoring the immense value of our investment in foundational data architecture,' said Vijay. The caution comes as regulatory frameworks like India's DPDP Act and the EU AI Act push CIOs to reevaluate AI risk, explainability, and accountability. 'Enterprises are navigating an evolving regulatory landscape (e.g., EU AI Act, India DPDP act, sectoral guidelines). Policy readiness is improving, with new corporate governance playbooks focused on responsible AI, but gaps remain—especially around explainability and real-time monitoring,' added Vijay. 'The tools are evolving rapidly. Culture and policy? That's where the gap lies. Many enterprises still have a command-and-control mindset, and introducing autonomous agents into that can be uncomfortable. There's a need to build digital trust, redesign workflows, and establish clear guardrails for autonomy to work responsibly,' Mehjabeen noted. 'For us, the most immediate opportunity is in internal operations and analytics. Think automated generation of daily reports, intelligent monitoring of IT infrastructure, or bots that can track anomalies and initiate alerts without human nudges. Over time, I see it expanding into customer-facing areas, but with tighter governance around decision-making boundaries', she added. Analysts say stakes are high—and so are the rewards According to McKinsey, Agentic AI could unlock $4.4 trillion in annual value globally, especially in industries like manufacturing and logistics. CIOs like Mehjabeen believe those gains are within reach. 'Absolutely. In fact, that's where I believe agentic AI could shine the brightest in industrial environments like ours. Imagine an agent monitoring sensor data in real-time, predicting a component failure, raising a purchase requisition, and even following up for approvals—all autonomously. That's not far-fetched anymore—it's where we're headed,' said Mehjabeen. Agentic AI is here—and it's powerful. But for Indian enterprises, the real challenge lies not in what the tech can do, but in what the organisation is ready to let it do.
Time of India
6 hours ago
- Time of India
Alphabet shares jump 3% as AI-driven spending fuels cloud revenue surge
Alphabet shares rose more than 3% in early trading on Thursday as the Google parent's earnings underscored a key message to investors: AI spending is climbing, but so are the returns. The tech giant has raised its 2025 capital spending forecast by $10 billion to $85 billion and signaled even higher outlay next year, stepping up efforts to meet soaring cloud demand and stay competitive in Silicon Valley's escalating AI race. Its cloud-computing unit delivered an almost 32% jump in second-quarter revenue, surpassing expectations, as investments in in-house chips and the Gemini AI model began to pay off. The results bode well for rivals Microsoft and both of which have been stepping up data center investments and operate larger cloud businesses. "Google came back fighting this quarter," said Bernstein analyst Mark Shmulik. "Investors have long been clamoring for Google to get more 'aggressive' in the AI race," he added. An early AI pioneer with its invention of the Transformer model - the foundation of most modern generative AI - Google appeared to fall behind OpenAI and Microsoft last year. But it has rebounded this year, with AI Mode reaching 100 million monthly users just two months into its wider rollout, and Gemini surpassing 450 million monthly users. Its ad business, which accounts for about three-quarters of its sales, also continues to fare well in the face of economic uncertainty wrought by tariffs and geopolitical tensions. Revenue in the business rose a better-than-expected 10.4%, a positive sign for rivals such as Meta and Snap that rely on digital ads for most of their revenue. At least 27 brokerages raised their price targets on Google stock after the results, taking the median target to $220 from $200 a month earlier. Still, some analysts warned the higher spending may draw fresh scrutiny from investors, who have largely stayed on the sidelines this year. Alphabet shares are up just 0.5% in 2025, trailing Microsoft's 20% increase and a 22% rise in Meta stock, also held back by regulatory battles that are looking to break its illegal monopoly in the search and the ad-tech markets. Alphabet's 12-month forward price-to-earnings ratio stands at 18.88, trailing Microsoft's 33.03 and Amazon's 33.31, according to data compiled by LSEG. "On paper, it has all the right tools to lead in AI - cutting-edge models and massive distribution," said Matt Britzman, senior equity analyst at Hargreaves Lansdown. "That said, until there's more confidence AI integration won't cannibalise core search revenue, and some clarity around ongoing legal battles, there's enough uncertainty to cap near-term upside."
