两会｜AI赋能产业发展存在哪些堵点痛点？

Core Viewpoint - The article discusses the urgent need for the integration of artificial intelligence (AI) with economic and social development in China, emphasizing the importance of computing power as a core infrastructure for AI advancement [2]. Group 1: Computing Power Development - Computing power should be categorized into training and inference computing. The demand for inference computing is expected to grow exponentially as the industry transitions into the "AI+" application era [3]. - There is a current gap in the supply of intelligent computing power that meets the demands of the times. It is suggested to establish an open platform for AI large model training computing power, dynamically allocating resources based on user needs [3]. Group 2: AI Technology Application - The transition from "computing power infrastructure" to "commercial closed-loop and governance collaboration" is critical. There is a tendency to focus on construction rather than application, leading to isolated innovations that fail to create scalable commercial value [4]. - Recommendations include accelerating the implementation of "AI + scenario closed-loop" demonstration projects in key areas like industrial manufacturing and smart finance, and establishing a governance system for AI [4]. Group 3: Talent Development and Data Governance - There is a need to reshape the talent cultivation system for the intelligent era, promoting educational reforms and establishing interdisciplinary programs to address the talent shortage in AI [5]. - A systematic approach to building an industrial data governance framework is essential to remove barriers for AI empowerment in manufacturing. This includes organizing technology breakthroughs and pilot demonstrations [5].