论人工智能的理解——以DeepSeek-R1等新兴人工智能为例  

On the Understanding of Artificial Intelligence:Taking Emerging AI such as DeepSeek-R1 as an Example

作　　者：李全兴 吴国林 LI Quanxing;WU Guolin(School of Marxism,South China University of Technology,Guangzhou 510641,Guangdong,China;Institute for Advanced Study in Science,South China University of Technology,Guangzhou 510641,Guangdong,China)

机构地区：[1]华南理工大学马克思主义学院,广东广州510641 [2]华南理工大学哲学与科技高等研究所,广东广州510641

出　　处：《华南理工大学学报(社会科学版)》2025年第2期9-18,共10页Journal of South China University of Technology(Social Science Edition)

基　　金：国家社会科学基金重大项目“当代量子诠释学研究”(19ZDA038);2020年度中宣部文化名家暨“四个一批”人才项目“技术诠释学研究”(x2sxN2200060)。

摘　　要：人工智能的理解标准为判断人工智能是否具有理解力提供了依据。人的理解是由认知能力、意志能力和解释能力共同作用而涌现出来的能力。而模仿于人的理解的人工智能的理解能力,是由认知能力与解释能力共同作用而彰显出来的。基于唐纳德·米奇对人工智能学习能力的分类,可将人工智能的理解划分为简单理解、具体理解与形式理解。简单理解能够对事物进行简单地识别或判断;具体理解能够运用相关概念并对具体事物进行归纳与推理;形式理解能对实在进行理解。而具备深度推理和解决复杂问题能力的DeepSeek-R1在一定程度上已初步具备形式理解能力。简单理解与具体理解属于认识论理解,形式理解属于本体论理解。人工智能的理解过程需要符合使用目的、客观事实等因素,这些因素是判断人工智能的理解是否合理的基本依据。The comprehension criterion for AI provides a basis for determining whether AI has comprehension.Human understanding emerges from the combined influence of cognitive ability,volitional ability,and interpretative ability.In contrast,the understanding ability of artificial intelligence,which mimics human understanding,is manifested through the interaction of cognitive ability and interpretative ability.Based on Donald Mitch's classification of AI's learning ability,AI's understanding can be divided into simple understanding,concrete understanding and formal understanding.Simple understanding can make simple identification or judgement of things;specific understanding can apply relevant concepts and make induction and reasoning about specific things;formal understanding can make sense of the real.DeepSeek-R1,which has the ability to reason deeply and solve complex problems,has been primed with formal understanding to some extent.Simple understanding and concrete understanding belong to epistemological understanding,and formal understanding belongs to ontological understanding.The process of understanding in artificial intelligence needs to align with factors such as intended use and objective facts.These factors serve as the fundamental basis for determining whether the understanding demonstrated by artificial intelligence is reasonable.

关 键 词：人工智能 理解标准 DeepSeek 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]