When AI goes wrong: Fatal errors in oncological research reviewing assistance Open AI based
Marwan Al‐Raeei
Abstract
In this letter to the editor, the use of artificial intelligence (AI) techniques, specifically the Chat-GPT based “Review Assistant” by Elsevier, for reviewing scientific articles is discussed. While the tool has many benefits such as detecting linguistic and typographical errors in manuscripts, it also has limitations. An example is highlighted where the AI gave an incorrect and potentially dangerous answer regarding the bond energies of molecules in an oral tumor. This mistake shows that the use of AI for scientific research evaluation can be a double-edged sword, as it may provide inaccurate information that could have serious consequences.
Topics & Concepts
MistakeSWORDComputer scienceArtificial intelligenceData scienceNatural language processingPsychologyWorld Wide WebLawPolitical scienceArtificial Intelligence in Healthcare and EducationRadiomics and Machine Learning in Medical ImagingMeta-analysis and systematic reviews