LMU research sheds light on the use of ChatGPT when writing research papers
6 Sept 2024
The study, led by LMU Visiting Professor Dr. Sergio Uribe and Dr. Ilze Maldupa from RSU, analyzed nearly 300,000 dental research abstracts published between 2018 and 2024.
The results showed that nearly 10 out of 100 papers contained these specific words, indicating a significant presence of AI-assisted writing in dental research. Some of these words are illustrated here. The dashed line marks the date ChatGPT was published.
Most important results:
The word "delve" saw the most significant increase in usage with a 17-fold increase in frequency following the release of ChatGPT;
Other words such as "transformative", "realm" and "revolutionize" also showed notable increases, indicating a shift in the use of language in dental research;
To validate their results, the researchers used dentist-specific terms such as "periodontal" and "caries" as negative controls and "COVID" as a positive control. The results showed a stable use of the negative controls and the expected increase and decrease in COVID-related terms, confirming the robustness of their methodology.
Prof. Dr. Sergio Uribe says: "The emergence of generative AI tools like ChatGPT will change the landscape of scientific publications. It used to be a challenge to write a paper, but that is no longer the case. As a result, the focus will shift less on writing a manuscript and more on the work itself. Also, open science practices such as pre-registration and publication of datasets will become even more important to enable the verification and replication of results. This will also increase the reproducibility of research. These open science practices are also crucial to prevent the spread of misinformation or disinformation, as has been the case during the COVID-19 pandemic, and to restore and strengthen public trust in science."
Generative AI offers a tremendous opportunity to promote open science practices. For example, LMU has pioneered the creation of the LMU Open Science Center(https://www.osc.uni-muenchen.de/) and the LMU Open Data Repository (https://data.ub.uni-muenchen.de/), which enables the university's researchers to publish their data according to the principles of open science. In addition, generative AI can help bridge the gap between researchers and the public by facilitating the creation of papers with less scientific jargon or adding summaries that are easily understood by the general public.
The researchers call for further investigation into the global impact of generative AI on academic writing and encourage the academic community to engage in discussions about the ethical use of these tools.
“LMU and LMU Dentistry are at the forefront of research into and use of AI, open data and methods to increase the reproducibility of science - topics that were also relevant at our WHO-ITU conference on AI in medicine and dentistry here at the hospital last month. AI and data-driven medicine have incredible potential to improve care for patients and the healthcare system. However, we need guidelines for the use of these new methods in practice, but also in research. This study makes an important contribution.”