&w=3840&q=100)
AI models of Google and OpenAI win milestone gold at global math contest
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.
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.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Hans India
15 minutes ago
- Hans India
Vidnoz AI Review 2025: The Ultimate Free Tool for AI-Powered Video Creation
Vidnoz AI is a fun and easy tool that helps people make videos using computer systems and smart voices. You do not now want a digicam or actors. Just type your story, and Vidnoz turns it into a cool video. Many humans use it in 2025 due to the fact it's free and rapid. This tool is top-notch for children, teachers, people, and small business owners. What Is Vidnoz AI? Vidnoz AI is a free website that helps you make movies using clever tools. It gives you more than 1,200 fake people referred to as avatars. These avatars can communicate, smile, and act like actual humans. You can even lead them to talk in over one hundred forty languages with the usage of more than 1,800 voices. You do not need to learn how to edit videos. Vidnoz offers you ready-to-use video styles, called templates. Just pick out one, write your message, and your video is completed in a single minute! It is one of the high-quality AI video generator free loose equipment you could locate online these days. Best Things About Vidnoz AI 1. Easy to Use Vidnoz is very simple. You just click on, type, and pick. There aren't any hard steps. Even a newbie could make a video rapidly. 2. Free to Try You could make up to ten motion pictures each day without paying. This allows you to analyze before you purchase anything. 3. Lots of Voices You can choose from glad voices, severe voices, or maybe make a duplicate of your very own voice. This is called voice cloning. 4. Different Kinds of People There are medical doctor avatars, cool animated film avatars, and commercial enterprise avatars. You can pick out the nice one in your story. Vidnoz even lets you change how avatars look—like using an AI clothes changer to match the outfit to your story idea. How People Use Vidnoz Teachers Teachers use it to train children in an amusing manner. For example, a technology lesson with a cartoon avatar is a greater laugh than studying an e-book. Marketers People who sell matters use it to reveal how their product works. It saves money due to the fact they do not now want a large video crew. Offices Offices make short movies to talk to their people. It is faster than sending lengthy emails. What's New in 2025? In 2025, Vidnoz introduced a laugh gear like Talking Photos : Make any picture communicate. : Make any picture communicate. AI Dance : Avatars can dance to the track. : Avatars can dance to the track. Real-Time Teamwork : Teams can paint collectively at the same video. : Teams can paint collectively at the same video. Google Voices: Over 1,000 new voices introduced. This new equipment makes it even more thrilling for folks who need creative and fun films. Things That Could Be Better Free films have a watermark (a small emblem). No 4K video yet, just handy HD (1080p). Some fancy equipment wants a paid plan. Still, most people love the loose plan and discover it is super useful. Why So Many People Love Vidnoz Over 500,000 people use Vidnoz as it works rapidly, saves cash, and is a laugh to apply. It is extraordinary for YouTubers, commercial enterprise groups, colleges, and everybody who wishes a cool video without hard paintings. It's one of the best tools while looking for an AI video generator unfastened with tons of cool functions. Conclusion Vidnoz AI is sort of a magic helper that makes motion pictures for you. It is best in case you want to keep time, make fun content, or train something in a groovy way. Try it these days for free, and you'll see how clean video-making may be!
Economic Times
15 minutes ago
- Economic Times
Has the pecking order changed for the IT majors? Kumar Rakesh explains
Kumar Rakesh, Associate Director- Equity Research, BNP Paribas, highlights the Indian IT sector's resilience over three decades. Despite challenges like the dotcom bubble and Covid, it has consistently grown. While Gen AI presents opportunities, its impact will vary across companies. Infosys and Persistent Systems are early leaders, leveraging partnerships and AI capabilities. ADVERTISEMENT With respect to Infosys, what are you building in and with respect to the commentary, which are going to be the key monitorables for you? Kumar Rakesh: For Infosys we are building relatively stronger growth. In the context of the rest of the IT services, we think Infosys will outperform the larger pack. We are building more than 2% sequential constant currency growth for the company this quarter, which will be partly helped by the acquisition they have done. About 40 bps of growth is coming in from that. But even excluding that, on an organic basis, we are expecting them to report a healthy 1.6-1.7% sequential growth and that is partly on a very low base of the last quarter. The fourth quarter for Infosys was quite weak and hence coming from a low base, it should see a pick-up in growth. Europe revival helps IT survive quarter shocks From the management commentary perspective, this earning season started with TCS results and sparked off the debate on whether the IT services demand has started seeing a material hit and the possibility of Indian IT services companies seeing a bigger impact compared to some of the larger peers because the commentary which was coming in from the global peers was far more stable demand environment. IT companies tighten belt as AI, macro headwinds squeeze biz margins Through the earning season we have seen that the commentary largely converges towards a stable demand environment. and we would look out for a similar commentary from Infosys as well, that the demand environment continues to be stable. Obviously the guidance as well built in the commentary and the guidance the company is going to give will be the key focus. What is the pecking order right now from the earnings that have come by? It was always TCS up there and then Infosys and then the rest? Has it changed at all after TCS and Wipro's earnings? Kumar Rakesh: At the start of this month as with our preview, we had shifted our preference to Infosys over TCS. Part of the reason was that we do see that TCS has some of the client specific issues that they will have a more lumpy performance. That being said, TCS valuation continues to look very attractive and hence there is a downside support in the TCS valuation. That being said, from an outperformance perspective, Infosys could be a preferred pick partly because of valuation, but more importantly because of the growth outperformance that the company can deliver compared to most of the other largecaps peers in the sector. The margin also seems to be relatively more stable. We will have to watch out how that goes out today as they report, but our expectation would be that Infosys earnings growth will be among the highest this year compared to most of the other largecaps. ADVERTISEMENT In terms of the verdict that we have seen across the board from largecap IT companies, so far we have either seen a miss versus estimates or inline, nothing positive. When do you see this trend reversing because of the guidance that some of the companies have given? It does say that by the end of FY26, they can expect some improvement in margins. When do you see this turnaround coming about? Kumar Rakesh: You are right as the general performance in this quarter has been quite mixed, largely a disappointment on growth as well as margin for many of the companies. A lot depends on the macro, how the US macroeconomic environment starts shaping up. There is a fair degree of uncertainty. A lot of tariff related deals are yet to be signed and that continues to elevate the uncertainty especially in verticals related to retail, CPG, and manufacturing, especially automotive within manufacturing and hence those verticals could take longer to start showing any recovery. The good news is that the BFSI vertical continues to remain quite resilient. Some of the companies even called out that even discretionary spending in that vertical seems to be there and hence that is one vertical which will continue to drive the growth for the next couple of quarters. ADVERTISEMENT Technology services is another vertical which seems to have bottomed out and started showing good results. These are the two verticals which we would expect to start driving growth and as the macro starts improving, the trade deal starts getting signed, and the uncertainty starts coming down, hopefully by the end of this financial year we would expect the rest of the verticals also to start showing recovery. Once that happens, the FY27 essentially becomes the recovery year which is where investors will start focusing on that this financial year will largely be a low single-digit growth year but can we get to the mid-single-digit sort of a growth in FY27 and if that happens, then the interest will start coming back into IT. ADVERTISEMENT The other debate that is going on is with respect to the AI transition because some experts believe that for Indian IT companies they have managed to get hold of the latest in technology and this can hold true yet again as well. Help us analyse your take on how Indian IT companies are expected to win this particular race. Do you believe that Indian companies are well versed to go ahead and adapt this particular technology and also which companies can actually be the early winners over here? Kumar Rakesh: We agree that if you go back over the last two-three decades starting from the mainframe transition, then the digital technologies came, a lot of SaaS players also came in the market. Through all those transitions the expectation was that the Indian IT services industry will take a hit but that did not happen. Over the last three decades, we have never seen Indian IT services industry's revenue to decline on a year-on-year basis, and so that is quite a resilient performance and during this period of time we have seen the dotcom bubble, the global financial crisis, multiple macroeconomic challenges in Europe and US and the Covid. Through all these phases, the Indian IT services industry has shown quite a strong resilience in its performance. We also believe it is a quite resilient industry. That being said, while we believe that on the Gen AI side, it would be net neutral to net positive for the industry, there would be some, who would be a bigger beneficiary and there could be some companies that could see a negative impact as well. It's a little too early to start calling out who would be negatively impacted but there are some companies who have started taking some initial lead in terms of building capability on Gen AI and also participating in clients' transformation on the Gen AI journey. Infosys is one of those companies. They have done a fair bit of work. Their partnership with some of the global Gen AI leading hardware companies as well as their own capability on the small language model which they are building seems to be quite promising. The other company which stands out is Persistent Systems that is doing a fair degree of work on Gen AI and some degree of that decoupling between the employee growth and revenue growth is already visible in that company. So, these are a couple of companies which have already started showing some initial signs of leadership outperformance from the Gen AI side and we are closely tracking that. ADVERTISEMENT (You can now subscribe to our ETMarkets WhatsApp channel)
Time of India
15 minutes ago
- Time of India
Mapping future: AI, electrification and automated testing drive auto tech forward
As the automotive world accelerates towards a smarter, cleaner and more connected future, this week's developments reveal a sector embracing complexity with innovation. From Yamaha's multi-pathway approach to electrification, to India's growing AI-powered EV ecosystem, and a global call for tighter oversight on autonomous technologies — the landscape is rapidly evolving. ETAuto brings you a curated round-up of key trends, technologies and transformations shaping mobility's next chapter. Electrification Is Just One Lane in Yamaha's Growth Roadmap Yamaha is exploring a diversified approach to growth in emerging markets, with electrification forming just one part of its broader strategy. Rather than a one-size-fits-all push, the company is balancing ICE, hybrid and EV options tailored to market needs. Read more Smarter Infrastructure: Maharashtra Launches Six Automated Vehicle Testing Stations Rosmerta Technologies has launched six automated vehicle inspection centres in Maharashtra to streamline and modernise vehicle compliance and roadworthiness checks, reinforcing the push toward digital governance in mobility. Read more AI in Motion: India's EV Ecosystem Attracts Smart Tech Players MediaTek is positioning itself as a key enabler of India's next-gen EVs, offering AI and high-performance computing platforms tailored to the country's unique needs. Meanwhile, Spyne has launched an AI assistant, VINNIE, designed to boost operational efficiency for used car dealerships. MediaTek targets Indian EV market with AI Spyne introduces VINNIE AI assistant Global Spotlight: Oversight Tightens on Self-Driving Tech In the US, a federal auto safety nominee has called for stricter oversight of autonomous driving systems, amid growing concerns around reliability and transparency. The move follows a Tesla driver's testimony that Autopilot failed to prevent a fatal crash. US auto safety nominee urges active oversight Tesla driver says Autopilot failed in fatal crash Mapping Smarter Roads: Genesys Integrates DIGIPIN in 2D & 3D Maps Genesys International has integrated DIGIPIN technology into its national mapping solutions, enhancing the precision and interactivity of 2D and 3D spatial data across India. This upgrade supports smarter urban planning, logistics, and autonomous navigation frameworks. Read more For insights into the fast-evolving automotive tech space, follow ETAuto for weekly analysis, trends, and deep dives. We'd love to hear what you think about this edition of the newsletter! Your feedback and suggestions help us improve and deliver content that matters to you.
