中国电化教育

生成式人工智能在嵌入拔尖创新人才培养同时，亦会产生相应风险。为此，该研究使用LDA主题模型、TF-IDF算法与深度访谈相结合的混合研究设计，发现现阶段生成式人工智能嵌入拔尖创新人才培养中会在培养主体、课程、评价以及管理层面产生相应风险。细究这些风险，其与生成式人工智能的模型、教育生态以及监管体系密切相关。为有效规避生成式人工智能嵌入拔尖创新人才培养的风险，应从“根源治理”“内部变革”“外部监管”方面入手，开发适配拔尖创新人才培养的智能模型，优化智能技术嵌入人才培养的教育生态与建立生成式人工智能应用的监管体系。

Generative artificial intelligence can empower the cultivation of top-notch innovative talents, but it will also generate corresponding risks. To this end, this study uses a hybrid research design that combines the LDA topic model, TF-IDF algorithm and in-depth interviews,and finds that the embedding of generative artificial intelligence in the cultivation of top-notch innovative talents at this stage will generate corresponding risks in the cultivation subjects, courses, evaluation and management levels. A detailed study of these risks is closely related to the model of generative artificial intelligence, the educational ecology and the regulatory system. In order to effectively avoid the risks of embedding generative artificial intelligence in the cultivation of top-notch innovative talents, we should start from “root governance”, “internal change” and “external supervision”, develop intelligent models suitable for the cultivation of top-notch innovative talents, optimize the educational ecology of embedding intelligent technology in talent cultivation and establish a regulatory system for the application of generative artificial intelligence.