OpenAI's Experimental AI Matches Gold Medal Math Olympiad Performance, GPT-5 Launch Soon
OpenAI researcher Alexander Wei announced the breakthrough on X, revealing that the experimental AI tackled five out of six problems from this year's IMO under authentic exam conditions. Scoring 35 out of 42 points, the model reached a level that would earn a human contestant a gold medal at the world's most challenging high school math competition.
'We evaluated our models on the 2025 IMO problems under the same rules as human contestants: two 4.5-hour exam sessions, no tools or internet, reading the official problem statements, and writing natural language proofs,' Wei stated.
The IMO is widely recognised for its notoriously demanding problems that test the deepest levels of mathematical creativity and reasoning. Wei noted that the AI's performance marks a significant leap forward compared to previous benchmarks. 'We've now progressed from GSM8K (~0.1 min for top humans) MATH benchmark (~1 min) AIME (~10 mins) IMO (~100 mins),' he added, illustrating the scale of advancement.
Independent grading by three former IMO medallists confirmed the model's solutions. The AI successfully solved problems P1 through P5 but did not complete P6. Wei made the model's detailed solutions public, pointing out its 'distinct style,' which reflects its experimental framework.
What sets this achievement apart is the AI's ability to generate complex, human-like proofs. 'By going beyond the reinforcement learning paradigm of clear-cut, verifiable rewards we've obtained a model that can craft intricate, watertight arguments at the level of human mathematicians,' Wei explained.
Despite this progress, OpenAI does not plan to release this specific IMO-level AI model to the public anytime soon. Wei clarified that while the company is gearing up for GPT-5's rollout, the Math Olympiad project remains separate and will continue behind closed doors. 'We don't plan to release a model with IMO gold level of capability for many months,' he added.
Echoing Wei's excitement, OpenAI CEO Sam Altman called the accomplishment 'a significant marker of how far AI has come over the past decade.' He emphasised that the IMO-level AI is not a narrowly trained math tool but part of broader research pushing general-purpose reasoning forward. 'We are releasing GPT-5 soon, but want to set accurate expectations: this is an experimental model that incorporates new research techniques. We don't plan to release a model with IMO gold level of capability for many months,' Altman reiterated.
Wei also looked back on his early forecasts, reflecting on how AI has surpassed expectations. 'In 2021, my PhD advisor, Jacob Steinhardt, had me forecast AI math progress by July 2025. I predicted 30 per cent on the MATH benchmark. Instead, we have IMO gold,' he wrote. He credited team members like Sheryl Hsu and Noam Brown for their contributions and congratulated this year's IMO competitors, noting that several OpenAI researchers were once IMO medallists themselves.
With GPT-5 on the horizon and experimental AI solving Olympiad-level math, OpenAI's latest strides are set to reshape what's possible for machine reasoning in the years ahead.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Time of India
2 hours ago
- Time of India
Microsoft's access to OpenAI tech is focus of contract talks
Synopsis The companies have discussed new terms that would let Microsoft use OpenAI's latest models and other technology even if the startup decides it has reached its goal of building a more powerful form of AI known as artificial general intelligence (AGI), according to two people familiar with the negotiations.
India Today
3 hours ago
- India Today
Sam Altmans says GPT 5 AI is so smart, it made him feel useless ahead of imminent launch
AI is evolving at an unprecedented pace and getting smarter and smarter with each passing day. This advancement is also bringing on the fear of AI and its potential to replace humans. While that day hasn't arrived yet, even OpenAI CEO Sam Altman admits he's unsettled by what's coming next. Sharing a recent incident involving the upcoming GPT-5, Altman reveals that the capabilities of this advanced model have left him feeling almost 'useless.'advertisementSpeaking on comedian Theo Von's This Past Weekend podcast, Altman revealed that when he recently used GPT-5 for some work, the model demonstrated such intelligence that it left him feeling 'useless relative to the intelligence of AI.' Altman shared that he had fed the AI a question he didn't fully understand:'I put it in the model (a question I got in an email I received) – this is GPT-5 – and it answered it perfectly. I felt useless relative to the AI in this thing that I felt I should have been able to do, and I couldn't, and it was really hard. But the AI just did it like that. It was a weird feeling.' Altman described the moment as a realisation of how powerful GPT-5 has Altman has already confirmed on X (formerly Twitter) that GPT-5 will be released 'soon.' While the company has not yet shared the exact dates, the teaser and the fact that Altman shared GPT-5 is already being used are fuelling speculation that the next-generation model is just weeks away. According to rumours, the advanced AI model from OpenAI could arrive as early as August, with OpenAI planning to release mini and nano versions of the model through its has already been spotted in limited tests, further fuelling speculation about its arrival and intensifying interest in what OpenAI has called its most advanced system yet. In fact, Altman had previously described the model as 'a system that integrates a lot of our technology,' with enhanced reasoning capabilities built in. The so-called 'o3 reasoning' engine—originally planned as a standalone system—will now be integrated directly into GPT-5, making it a more unified and powerful on the same podcast, Altman also touched upon the raised concerns about the mental health impact of AI, particularly as more people are using AI for therapy, forming emotional attachments to AI companions, and relying on them extensively. Answering Von's question on whether AI will have the same negative effect that social media really has, Altman replies, 'I'm scared of that...I don't think we know quite the ways in which it's going to have those negative impacts, but I feel for sure it's going to have some and we'll have to". - Ends
Time of India
3 hours ago
- Time of India
Microsoft's AI edge under scrutiny as OpenAI turns to rivals for cloud services
Academy Empower your mind, elevate your skills Microsoft investors head into Wednesday's earnings with one big question: is the company's artificial intelligence edge at risk as partner OpenAI turns to rivals Google, Oracle and CoreWeave for cloud services?Exclusive licensing deals and access to OpenAI's cutting-edge models have made Microsoft one of the biggest winners of the generative AI boom , fueling growth in its Azure cloud business and pushing its market value toward $4 the April-June quarter, the tie-up is expected to have driven a 34.8% increase in Azure revenue, in line with the company's forecast and higher than the 33% rise in the previous three months, according to data from Visible that deal is being renegotiated as OpenAI eyes a public listing, with media reports suggesting a deadlock over how much access Microsoft will retain to ChatGPT maker's technology and its stake if OpenAI converts into a public-benefit conversion cannot proceed without Microsoft's sign-off and is crucial for a $40 billion funding round led by Japanese conglomerate SoftBank Group, $20 billion of which is contingent on the restructuring being completed by the end of the which recently deepened its Oracle tie-up with a planned 4.5 gigawatts data center capacity, has also added Google Cloud among its suppliers for computing analysts said investor views on the Microsoft-OpenAI partnership are divided, though the software giant holds an upper hand. "Microsoft's leadership earned enough credibility ... such that the company will end up negotiating terms that will be in the interest of its shareholders," the analysts of that confidence is reflected in the company's stock price, which has risen by more than a fifth so far this the April-June period, Microsoft's fiscal fourth quarter, the company likely benefited from a weaker dollar, stronger non-AI Azure demand and PC makers pulling forward orders for its Windows products ahead of possible U.S. is expected to have risen 14% to $73.81 billion, according to data compiled by LSEG, its best growth in three is estimated to have increased 14.2% to $25.16 billion, slightly slower than the previous quarter as operating costs spending will also be in focus after rival Alphabet raised its annual outlay by $10 billion last week. Microsoft has repeatedly said it remains capacity constrained on AI, and in April signaled continued growth in capex after planned spending of over $80 billion last fiscal year, though at a slower pace and on shorter-lived assets such as AI Morgan, senior portfolio manager at Synovus Trust who owns Microsoft shares, said the spending has been paying off."Investors may still be underestimating the potential for Microsoft's AI business to drive durable consumption growth in the agentic AI era."
