AI Accelerates Creation of High-Quality Medical Reviews, But Experts Warn Against Overconfidence

Researchers have successfully employed artificial intelligence (AI) to reproduce 12 systematic reviews in just two days, significantly accelerating a process that typically takes months. This breakthrough, detailed in a preprint on medRxiv, utilizes large language models (LLMs) to streamline key steps in the systematic review process, including updating existing reviews in as little as 20 minutes. The innovation, developed by a team including Christian Cao from the University of Toronto, has the potential to influence medical practice guidelines and policymaking. However, skepticism remains among some researchers who argue that essential tasks have yet to be automated, and independent verification of the findings is necessary. Systematic reviews are vital as they synthesize results from various studies, aiding decisions in medicine and policy. While AI tools have been used previously, the recent surge of interest highlights ongoing debates about their effectiveness and reliability in the field.

Source link