AI in the ER: What Is Watson and What Does It Mean for Medical Librarianship?
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
It runs on ten server racks, has natural language analysis capability and recently won a Jeopardy! Tournament where it was pitted against two champions. If you answered, “What is Watson”, you would be correct, but you probably didn’t answer in time to beat the new champ. Watson, the latest incarnation of IBM’s Deep Blue research project, recently astonished the Jeopardy! audience with its performance. This was a watershed moment in artificial intelligenceresearch because it required not only analytical reasoning skills (of the same sort that enabled IBM’s Deep Blue to become a chess champion), but also natural language processing skills. Simply gathering data and organizing it is nothing new for a computer system, but being able to apply a solid level of semantic reasoning to complex natural language questions is a much more challenging goal. By coupling sophisticated natural language recognition with the brute power of modern digital computing, Watson was able to handily defeat human opponents.