OpenAI’s LLM just earned a gold medal at the 2025 International Mathematical Olympiad (IMO)—a milestone that many AI researchers once thought was a decade away. The experimental reasoning model, run under the same 4½‑hour, no‑internet rules as human contestants, solved five of the six problems and scored 35 out of 42 points, comfortably above this year’s gold cut‑off. Three former […]| NATURAL 20
From Voice Command to Action: How Comet Works Perplexity’s Comet layers a voice‑driven agent atop a Chrome‑compatible browser. After a quick OAuth handshake, it gains scoped access to Gmail, Calendar, retail sites, and social networks. You speak or type a natural‑language command—“unsubscribe from Grubhub,” “add Taco Tuesday at eleven,” “find the cheapest Casa Azul”––and Comet orchestrates […]| NATURAL 20
The Breakthrough Claim A new preprint from a Chinese research team argues that artificial intelligence may have crossed a milestone: autonomous improvement of its own architecture. Their framework—dubbed ASI Arch—purports to design, test, and iterate neural‑network layouts without any human engineering. The authors liken the moment to AlphaGo’s self‑play leap in 2016, suggesting that AI research […]| NATURAL 20
Why an AI Playbook Now? Generative models are already drafting legal briefs, diagnosing disease, and steering factory robots. U.S. leaders fear that moving too slowly will cede long‑term advantages to geopolitical rivals, yet shipping untested systems could spark economic upheaval or erode public trust. The White House’s new national action plan tackles this dilemma by […]| NATURAL 20