生成式人工智能赋能社区智慧养老模式构建路径研究

Generative AI-Empowered Community Smart Elderly Care: A Study on Model Construction Path

摘要: 人口老龄化与养老服务供给不足之间的结构性矛盾，亟需通过技术创新重构社区养老模式。基于TOE模型，系统探讨生成式人工智能从技术、组织、环境三个维度赋能社区智慧养老模式构建，但在实践落地中面临AI“幻觉”引发的安全风险、人机协同责任边界模糊以及基础设施与资源配置不均等现实困境。针对上述困境，应从风险防控层面构建垂直领域大模型与人工复核机制、从权责明晰层面确立分层担责与全程追溯框架、从资源普惠层面推进分级建设与区域协同共享，以推动生成式人工智能赋能社区智慧养老从理论构想走向实践落地。

Abstract: The structural contradiction between population aging and insufficient supply of elderly care services urgently requires the reconstruction of community elderly care models through technological innovation. Based on the TOE framework, this study systematically explores how generative artificial intelligence empowers the construction of community smart elderly care models from three dimensions: technology, organization, and environment. However, in practical implementation, it faces realistic dilemmas including safety risks caused by AI “hallucination,” blurred responsibility boundaries in human-machine collaboration, and uneven infrastructure and resource allocation. To address these dilemmas, countermeasures should be taken from three aspects: building vertical domain-specific large models and establishing manual review mechanisms for risk prevention, establishing a hierarchical responsibility-sharing and full-process traceability framework for clarifying rights and responsibilities, and promoting tiered construction and regional collaborative sharing for resource universal benefit. These efforts aim to advance generative artificial intelligence-empowered community smart elderly care from theoretical conception to practical implementation.