In a groundbreaking display of AI advancement, Google and OpenAI showcased their latest models at the 2025 International Collegiate Programming Contest (ICPC) in Baku, Azerbaijan. Google’s Gemini 2.5 Deep Think achieved a gold-medal score by solving 10 out of 12 complex algorithmic problems, including a human-unsolvable “multi-dimensional optimization” puzzle. In a remarkable counteraction, OpenAI’s GPT-5 secured a perfect score by tackling all 12 problems flawlessly, indicating profound capabilities in multi-step and abstract reasoning. The ICPC, renowned as the premier platform for university-level coding talent, featured 139 elite teams competing to solve intricate challenges in a five-hour timeframe. This rivalry not only highlights advancements in AI’s problem-solving prowess but also signals a new epoch in human-AI collaboration, where AI serves as a robust partner in complex problem-solving across various domains. Both companies’ achievements underscore their commitment to pushing boundaries in artificial intelligence and algorithmic mastery.
Source link
