Recent demonstrations of artificial intelligence capabilities have highlighted remarkable advances in mathematical problem-solving and strategic reasoning. At a confidential event in Berkeley, 30 leading mathematicians attempted to challenge OpenAI's o4-mini model with unsolved mathematical problems, but the AI successfully solved most of them within minutes, showcasing advanced reasoning skills. Parallel developments in China reveal that domestic AI models have significantly improved their performance on the national college entrance examination (Gaokao) mathematics test, with scores rising from 47 to over 130 points in one year, surpassing the average human level. Notably, models such as Doubao and DeepSeek achieved high marks in both multiple-choice and problem-solving sections, while Google's Gemini led in objective question performance. These advancements suggest that AI is rapidly approaching a level of mathematical proficiency that could position it as a powerful collaborator in research and education. Additionally, AI models have recently competed in complex strategic games like Diplomacy, further demonstrating their growing cognitive capabilities.