Workflow
企业智能搜索服务
icon
Search documents
所有人都在谈“人工智能+”,到底怎么落地?
腾讯研究院· 2025-09-02 08:23
Core Viewpoint - The article discusses the transition from "Internet+" to "Artificial Intelligence+" as a new phase in technological integration, emphasizing the transformative potential of AI in reshaping industries and societal operations [5]. Group 1: Differences Between "Artificial Intelligence+" and "Internet+" - The technological stage differs, with "Internet+" being based on mature digital technologies while "Artificial Intelligence+" is characterized by rapid iteration and uncertainty in technology and applications [7]. - The value creation mechanism varies; "Internet+" enhances connectivity, while "Artificial Intelligence+" focuses on computational enhancement, improving productivity at each node and expanding the network's value [10]. - The diffusion paths are distinct; "Internet+" follows a consumer-to-producer model, while "Artificial Intelligence+" is more producer-focused, requiring deep integration into business processes before reaching consumers [12]. Group 2: Economic Impact of AI - AI's productivity effects are expected to grow exponentially, with predictions that AI could contribute to a 15% increase in global economic growth over the next decade [11]. - The rapid evolution of AI capabilities, with task completion abilities doubling approximately every seven months, indicates a significant potential for economic value creation [11]. Group 3: Practical Exploration of "Artificial Intelligence+" - Companies should prioritize high-value AI use cases that are data-rich and core to their business, as demonstrated by Pfizer's use of AI to enhance drug development efficiency [17]. - The engineering of AI systems is crucial, with companies needing to adapt general models to specific business needs through techniques like prompt engineering and retrieval-augmented generation [18]. - Building AI datasets should focus on business needs rather than data collection for its own sake, ensuring that data strategies are integrated throughout the AI application lifecycle [19]. Group 4: Recommendations for Promoting "Artificial Intelligence+" - A top-level design is necessary to create an innovative environment for "Artificial Intelligence+", similar to the strategic guidance that supported "Internet+" [22]. - Encouraging a diverse range of developers and startups in AI applications can foster innovation and investment in the sector [23]. - Establishing a comprehensive data element market and promoting open industry application scenarios can enhance the sustainable development of AI applications [25].