Can artificial intelligence (AI) alleviate the laborious and time-intensive nature of academic research collection? An international team of researchers delved into the credibility and efficiency of generative AI as an information-gathering tool in the medical field.
Led by Professor Masaru Enomoto from the Graduate School of Medicine at Osaka Metropolitan University, the research team conducted an experiment where identical clinical questions and literature selection criteria were presented to two generative AIs: ChatGPT and Elicit. The findings revealed a stark contrast — while ChatGPT suggested fictitious articles, Elicit demonstrated efficiency by providing multiple references within a few minutes, maintaining the same level of accuracy as the researchers.
Dr. Enomoto shared insights into the motivation behind the research, stating, “This study originated from our challenges in managing extensive volumes of medical literature over prolonged periods. Although caution is advised due to the nascent stage of generative AI information access, as current information may lack accuracy and currency, the evolution of ChatGPT and similar AIs holds promise to revolutionize the landscape of medical research in the future.”
As AI continues to advance, the intersection of technology and academia remains a dynamic space, with ongoing developments shaping the potential impact of these intelligent systems on the future of medical research.
