中欧人工智能存在相向而行的发展空间
Di Yi Cai Jing·2025-08-17 11:29

Core Insights - The article discusses the potential collaboration between the European Union (EU) and China in the field of artificial intelligence (AI), highlighting their differing approaches and the opportunities for shared technological advancements and standards [1] Innovation Layer: Complementary Cooperation - Europe has world-class research institutions that provide a solid foundation for AI model technology, but it lacks the computational resources compared to the US, which holds 75% of global AI computing power [2] - The EU plans to invest €30 billion in AI infrastructure to enhance its innovation ecosystem, including the establishment of 13 regional AI factories across 17 member states [2] - The EU faces challenges in energy supply for its data centers, which require significant power, and the lack of a unified fiscal policy complicates funding for the €30 billion investment [3] Application Layer: Expanding Boundaries Together - Global AI development has shifted from experimental innovation to practical applications, with the EU focusing on vertical sectors like healthcare and agriculture due to limited computational resources [5] - China's AI models have advanced beyond the EU's specialized models, with a strong presence in various application scenarios, and the country is rapidly expanding its AI capabilities [6] - Collaborative opportunities exist in AI applications, with both regions able to learn from each other's strengths and experiences [7] Governance Layer: Seeking Common Ground in Regulation - The EU is leading in AI governance with its comprehensive AI regulation, which categorizes AI systems by risk levels and imposes strict compliance requirements [8] - The EU aims to establish itself as a global leader in AI governance, while the US is pursuing its own path to dominate AI governance through technological supremacy [9] - There is a window of opportunity for collaboration between the EU and China in AI governance, particularly in areas like risk classification and human control principles [10][11]