Download |

ABSTRACT

Artificial intelligence (AI) is powerful, complex, ubiquitous, often opaque, sometimes invisible, and increasingly consequential in our everyday lives. Navigating the effects of AI as well as utilizing it in a responsible way requires a level of awareness, understanding, and skill that is not provided by current digital literacy or information literacy regimes. Algorithmic literacy addresses these gaps. In arguing for a role for libraries in algorithmic literacy, the authors provide a working definition, a pressing need, a pedagogical strategy, and two specific contributions that are unique to libraries.

INTRODUCTION

Algorithms, in one form or another, are as old as human problem solving and as simple as “a sequence of computational steps that transform the input into the output.”1 For centuries they have been effective, and uncontroversial, methodologies. However, the rise of artificial intelligence (the integration of big data, enhanced computation, and advanced algorithms) with its human and greater-than-human performance in many areas has positioned algorithms as transformational and a “major human rights issue in the twenty-first century.”2

Algorithmic literacy is important given of the prevalence of algorithmic decision-making in many aspects of everyday life and because “the danger is not so much in delegating cognitive tasks, but in distancing ourselves from—or in not knowing about—the nature and precise mechanisms of that delegation.”3 As a result, David Lankes warns of a new type of digital divide with “a class of people who can use algorithms and a class used by algorithms.”4 In a 2019 Deloitte survey “only 4percent reported they were confident explaining what AI is and how it works.”5 While a 2019 Edelman survey indicated general awareness of AI, it also revealed a similar lack of knowledge about the details of AI.6 An informed, algorithmically literate public is better able to negotiate and employ the complexities of AI.7

Identifying and acting upon algorithms as a literacy makes them as “fundamental as reading, writing, and arithmetic.”8 However, the uncritical use of the term literacy should make one suspicious of extending it to algorithms. Increasingly “literacy” has come to mean merely a body of knowledge or a set of domain-specific skills.9 Various literacies have been described, such as health, death, financial, physical, ocean, religious, visual, dancing, spatial, screen, and porn. This includes a dozen different technology-related literacies.10 The case for algorithmic literacy, and the role for libraries in advancing it, must rest on a clear definition, a recognized problem and need, a pedagogical strategy, and a unique (or at least supportive) contribution libraries can provide.

Michael Ridley (mridley@uoguelph.ca) is Librarian Emeritus, McLaughlin Library, University of Guelph, Ontario, Canada. Danica Pawlick-Potts (dpawlic@uwo.ca) is PhD Candidate, Faculty of Information and Media Studies, Western University, Ontario, Canada. © 2021