深度｜吴恩达：中国在开源权重模型的发布方面已经远远领先于美国；很多人用Agentic AI的方式是错的

Core Insights - The conversation highlights the current state of artificial intelligence (AI), emphasizing the importance of understanding and utilizing AI effectively in various fields [3][4][11] - Andrew Ng stresses the need for the U.S. to maintain its competitive edge in the global AI race, particularly in the context of open-source models and talent acquisition [20][22][30] Group 1: AI Development and Trends - AI's capabilities are evolving, with a focus on iterative workflows rather than simple prompt-based outputs, which can enhance the quality of tasks such as medical and legal advice [12][14] - China has taken a lead in releasing open-source weight models, which are freely available for global use, surpassing the U.S. in cumulative adoption [7][23] - The concept of "Agentic" systems was introduced to streamline discussions around AI capabilities, moving beyond binary debates on whether a system qualifies as an agent [9][10][11] Group 2: Education and Skills for the Future - Understanding programming and computer languages will remain crucial, as those who possess these skills will have a significant advantage in the job market [15][16] - The current educational system must adapt to ensure that future generations are equipped with the skills to create software, not just use it [16][18] Group 3: Policy and Global Competition - Concerns are raised about U.S. immigration policies affecting the influx of talent, which is vital for maintaining competitiveness in AI [20] - The need for a balanced approach to AI regulation is emphasized, advocating for innovation rather than restrictive measures that could stifle growth [22][30] Group 4: Future of AI and Public Perception - There is a significant public distrust towards AI, which needs to be addressed through clear communication about its benefits and potential [24][25] - The importance of making AI tools accessible and providing training to enhance public trust and utilization is highlighted [25][30]