A commentary coauthored by Assistant Professor Xu Ying of the School of Government at the Beijing-based Central University of Finance and Economics, or CUFE, was published on Aug 11 in the correspondence section of the prestigious international scientific journal Nature. The article was entitled AI tools speed up analysis, but scientific truths must be grounded in reality. It explored the application of artificial intelligence in research and its implications for scientific authenticity.

The commentary made the point that while AI can boost research efficiency, it might also lead to an over-reliance on automated modeling tools. Complex models can only simplify simulations of environmental systems, but they cannot replace the understanding of researchers about real-world geographical environments. The output of computer simulations still requires validation through field observation data, it asserted. Although AI broadens the scope and dimensions of data analysis in research, it cannot substitute for rigorous empirical investigations. Human field observations and on-the-ground assessments should remain central to identifying objective environmental patterns and assessing the reliability of research conclusions.

AI is now deeply integrated into the systems of industry, academia and research, as well as throughout the educational continuum from undergraduate to doctoral levels — encompassing key activities such as literature reviews, data analyses and research training. As institutions of higher learning develop AI application skills, they must be cautious about the situation in which researchers and students would become overly dependent on these tools. They should reinforce training in core skills such as original data collection, independent thinking, scientific judgment, and empirical validation. Researchers should be guided to understand the scientific questions underlying models, and to cultivate the ability to assess the scientific validity and credibility of research findings.

Reviewers: JIANG Ling

Editors: WANG Xinyu

Approvers: HUO Xiaoran