Southeast Asia Information Port (www.dnyxxg.com) – On the 13th, reporters learned from the Shanghai Municipal Commission of Economy and Informatization that Shanghai is comprehensively implementing the "Artificial Intelligence+" initiative to accelerate the creation of a new intelligent economy.
Shanghai's "Artificial Intelligence+" initiative currently focuses on three main areas: manufacturing, producer services, and AI4S (Artificial Intelligence for Science, AI-driven scientific research).
Shanghai is comprehensively implementing the "AI + Manufacturing" initiative through a "one industry, one policy" approach. For example, it supports Shanghai Electric in building an AI pilot-scale base for high-end equipment.
Shanghai supports AI in empowering producer services, focusing on process-oriented and knowledge-intensive industries such as law, accounting, finance, and human resources. Through AI, a new model of "intelligent agent execution + expert decision-making" is being formed, better transforming technological advantages into cost advantages.
Regarding AI4S, Shanghai is focusing on supporting 100 teams to advance 100 projects, supporting developers in fields such as biomedicine, new materials, and basic science to share scientific large-scale models and other toolchains.
On the other hand, facing the risks and challenges of AI information pollution, AI security cannot be ignored.
Turing Award laureate and Chinese Academy of Sciences Academician Yao Qizhi stated at the 2nd Pujiang AI Academic Conference in Shanghai that security is a major bottleneck in AI development. Without secure robots and AI systems capable of handling critical tasks, their applications will not be widely accepted.
Focusing on the intelligent transformation of industries with high security requirements, the Shanghai Artificial Intelligence Laboratory recently released the Intern-Shannon intelligent agent operating system. Hu Xia, Assistant Director and Leading Scientist of the Shanghai Artificial Intelligence Laboratory, pointed out that focusing on security is not about "putting the brakes" on AI, but rather about creating "natively secure AI" to allow the technology to continue its rapid development.
Chen Mingang, Deputy Director of the Artificial Intelligence Governance Institute of the Shanghai Computer Software Technology Development Center, revealed that the group standard "Intelligent Agent Evaluation Indicators and Methods," with multi-party participation in Shanghai, is currently undergoing internal testing. He believes that dedicated intelligent agents with clear boundaries and limited functions have relatively controllable risks and can be implemented first; general-purpose intelligent agents have less clearly defined functional boundaries and are difficult to comprehensively evaluate. (End)
