| 摘要: |
| 开源大模型为推动中医药教育数字化发展提供了重要契机。以DeepSeek为例进行探讨,DeepSeek展现出模型推理效率的深度优化、技术开放设定的开源创新以及公布思维链的推理过程的鲜明优势,在中医药教育中的主要应用包括医学知识传授模式重塑、临床实践技能训练优化以及中医药教育评价体系构建等方面。然而,DeepSeek类开源大模型在赋能中医药教育的过程中也会产生相应问题,存在“价值偏向”“认知固化”以及“智能幻觉”的潜在风险,进而影响DeepSeek赋能中医药教育的实际效果。因此,在实践中应从以价值引领赋能、以素养驾驭赋能、以技术驱动赋能3方面加以规避。 |
| 关键词: DeepSeek 中医药教育 大模型 开源 |
| DOI:10.13506/j.cnki.jpr.2026.01.017 |
| 基金项目:黑龙江省高等教育教学改革项目思想政治理论课教学改革研究专项(No.SJGSZ2022003) |
|
Publication Date:January 26, 2026
|
|
| DeepSeek empowers traditional Chinese medicine education:Applications,risks,and countermeasures |
| YU Qinming1,WANG Qifan2,GUO Hongwei1 |
| (School of Marxism, Heilongjiang University of Chinese Medicine, Harbin 150040, China) |
| Abstract: |
| Open-source large models provide a significant opportunity for advancing the digital transformation of medical education.Taking DeepSeek as an example, this study explores its distinctive advantages, including deep optimization of model reasoning efficiency, open-source innovation in technical settings, and transparent reasoning processes with disclosed chain-of-thought. Its primary applications in medical education encompass reshaping knowledge delivery methods, optimizing clinical skill training, and constructing evaluation systems. However, open-source large models like DeepSeek also introduce challenges during their integration into medical education, such as potential risks of "value bias," "cognitive rigidity," and "AI hallucination," which may undermine their practical effectiveness. Therefore, countermeasures should be implemented in practice, focusing on value-driven empowerment, competency-based adaptation, and technology-enhanced solutions. |
| Key words: DeepSeek Medical education Large language models (LLMs) Open-source |