A study focusing on the sources of information within the field of environmental science has shown ChatGPT perpetuates the Matthew Effect in science, where the rich get richer in terms of citations. It tends to cite highly-cited publications and exhibits a preference for older publications, and predominantly refers to well-respected journals in the field, with Nature being the most cited journal by GPT. Interestingly, GPT seems to exclusively rely on citation count data from Google Scholar for the works it cites, rather than utilizing citation information from other scientific databases such as Web of Science or Scopus.

The author Eduard Petiska warns that this can lead to a lack of diversity in the field, as certain researchers and institutions become more dominant and others are marginalized. This can limit the range of perspectives and approaches that are considered in environmental science research. It can also an perpetuate existing power imbalances, as those who are already in positions of power and influence continue to accumulate more resources and recognition. This can make it difficult for new voices and perspectives to be heard and taken seriously in the field.

The study (12pages) is here: https://arxiv.org/abs/2304.06794