Advertising on the Telegram channel «Artificial Intelligence | AI News»

🥇 OpenAI grabs GOLD in IMO 2025 OpenAI grabs GOLD in IMO 2025 with latest experimental reasoning LLM breaking new ground in general-purpose reinforcement learning and test-time compute scaling. Why this is a HUGE deal: - World’s most prestigious math competition—the International Math Olympiad (IMO). - A new level of sustained creative thinking compared to past benchmarks. - Reasoning time from GSM8K (~0.1 min for top humans) → MATH benchmark (~1 min) → AIME (~10 mins) → IMO (~100 mins). - Going beyond the RL paradigm of clear-cut, verifiable rewards to craft intricate, watertight arguments at the level of human mathematicians. - The model solved 5 of the 6 problems on the 2025 IMO. For each problem, three former IMO medalists independently graded the model’s submitted proof, with scores finalized after unanimous consensus. The model earned 35/42 points in total, enough for gold! Congratulations to the entire team. Sam Altman was right this a hot AI summer and team is busy building the future of AI. Please note this model will not be released for many months. @aipost 🪙 | Our X 🥇

Elon Musk on X. @aipost 🪙 | Our X 🥇

Scientists from OpenAl, Google DeepMind, Anthropic and Meta have abandoned their fierce corporate rivalry to issue a joint warning about Al safety. More than 40 researchers published a research paper today arguing that a brief window to monitor Al reasoning could close forever - and soon. @aipost 🪙 | Our X 🥇

OpenAI new model o3 alpha is a beast at front end coding - far better than Sonnet, o3, Gemini 2.5 Pro, or Grok 4. It created this UNBELIEVABLE procedurally generated planet with Three js. See the comparison with renders from other models. Get ready for a showdown with DeepSeek R1 and Kimi K2... @aipost 🪙 | Our X 🥇

Eric Schmidt asks what happens when AI creates millions of polymaths History's greatest discoveries came from experts applying ideas from one field to another — something AI can't do yet. "The rules keep changing". Solving non-stationarity might bring breakthroughs, but it'll take years to know @aipost 🪙 | Our X 🥇

Eric Schmidt says we're on the verge of AI replacing most programming and math tasks These domains are scale-free and don't require real-world data — just compute. "World-class AI mathematicians will emerge in the next year, and top-tier AI programmers within two" @aipost 🪙 | Our X 🥇

🔒 Universal Cancer Vaccine Nears Human Trials Scientists are developing a universal cancer vaccine that trains the immune system to recognize and destroy cancer cells across multiple tumor types. Early animal studies show remarkable results in shrinking tumors and preventing relapse. This innovation could transform cancer treatment, offering a single, adaptable therapy against countless cancers. It reflects AI’s growing role in accelerating drug discovery and brings us closer to a future where cancer might become a manageable disease. @aipost 🪙 | Our X 🥇

🔔 DeepMind leader calls OpenAI’s IMO gold claim bullshit. “IMO has an internal marking guide no one outside sees. Without that, you can’t claim a medal. With the point you lose on P6, you’re Silver, not Gold.” Reminder: > OpenAI didn’t even collaborate with IMO. > vague-posts results with no full transparency > DeepMind collaborated with IMO to verify results > IMO explicitly requested AI labs to wait a week after the closing ceremony > to avoid stealing spotlight from brilliant students (humans) > OpenAI chose to announce their results BEFORE the ceremony even ended > Research community: “OpenAI has been disrespectful” @aipost 🪙 | Our X 🥇

That's just the beginning. Other billionaires interested in AI will follow. Wondering who will build the first 10 GW, and then the 100 GW, data center @aipost 🪙 | Our X 🥇

Geoffrey Hinton says LLMs aren't just mimicking, they're making creative connections When asked why a compost heap is like an atom bomb? GPT-4 pointed to chain reactions, which showed it understood abstract links. "It saw the analogy during training, not test time". So, they're seeing analogies to compress knowledge, like humans do @aipost 🪙 | Our X 🥇