Mengaldo, Gianmarco. “Explain the Black Box for the Sake of Science: The Scientific Method in the Era of Generative Artificial Intelligence.” The EGU Interactive Community Platform, EGUsphere Preprint Repository, 25 July 2025, egusphere.copernicus.org/preprints/2025/egusphere-2025-3509/egusphere-2025-3509.pdf. Accessed 29 Oct. 2025.
Mengaldo’s piece describes the importance of the scientific method in research, and that for AI to be effectively used within the sciences, it too must be subject to the same principles. “…AI can be leveraged for scientific discovery via explainable AI”. New scientific knowledge might be possible to develop from the use of generative AI, but Mengaldo stresses the necessity for there to be interpretability in how AI reaches the conclusions that it does; the elucidation of the black box.
This piece addresses one of my greatest concerns with the use of AI in the sciences. While typically a field I would be somewhat more willing to permit a degree of usage, there is still the significant risk of bad training data resulting in bad results. That same issue arises when bad conclusions stem from bad assessments, but if we cannot know the process by which a conclusion is reached, we cannot allow ourselves to trust the conclusion itself.
Leave a Reply