Teaching CRAAP to Robots: Artificial Intelligence, False Binaries, and Implications for Information Literacy

Abstract

Researchers studying artificial intelligence and semantic computing are developing algorithms capable of processing large amounts of textual data and rendering judgment on its contents. Specifically, the field of sentiment analysis is focused on creating code that applies what programmers call “common sense” to evaluate whether writing is factual or opinionated, as well as how emotional the author was. This presentation will argue that these algorithms rely on false binaries, over-simplification, and poorly-constructed checklists, similar to the approach often used when discussing information literacy with first-year college students. Instead of employing this approach, this session will argue that librarians must recognize that human interpretation lies at the core of information literacy, and that we need to embrace that complexity rather than depend on algorithmic evaluation.