Pinski, Marc (2024)
Artificial Intelligence Literacy - Conceptualization, Measurement, Enablement, and Its Impact on Individuals and Organizations.
Technische Universität Darmstadt
doi: 10.26083/tuprints-00028092
Ph.D. Thesis, Primary publication, Publisher's Version
Text
Dissertation_Marc_Pinski.pdf Copyright Information: In Copyright. Download (14MB) |
Item Type: | Ph.D. Thesis | ||||
---|---|---|---|---|---|
Type of entry: | Primary publication | ||||
Title: | Artificial Intelligence Literacy - Conceptualization, Measurement, Enablement, and Its Impact on Individuals and Organizations | ||||
Language: | English | ||||
Referees: | Benlian, Prof. Dr. Alexander ; Jussupow, Prof. Dr. Ekaterina | ||||
Date: | 24 September 2024 | ||||
Place of Publication: | Darmstadt | ||||
Collation: | XXI, 235 Seiten | ||||
Date of oral examination: | 13 September 2024 | ||||
DOI: | 10.26083/tuprints-00028092 | ||||
Abstract: | Advancements in technology continually redefine what it means to be technologically literate in contemporary society and the business world. The recent surge in artificial intelligence (AI) technologies has particularly catalyzed this transformation, necessitating a reevaluation of existing technology literacy concepts. While AI technologies have achieved astonishing capabilities, they also have unique facets that distinguish them from prior technology, such as their inscrutability. These dynamics have prompted researchers in Information Systems (IS) and related disciplines to delve into the topic of AI literacy, investigating what it means to be literate concerning this new class of technologies. AI literacy refers to a holistic human proficiency in a variety of subject areas concerning AI that enable purposeful, efficient, and ethical usage of AI technologies. However, our current understanding of this new form of literacy is still quite limited. Since AI as a phenomenon is not only technologically different from prior technology but also has distinct sociological and psychological effects on humans, it remains unclear what such a new AI literacy concept needs to entail to prepare humans for the efficient and responsible usage and management of these technologies. Moreover, little research exists on the specific effects of AI literacy on humans and organizations, which is crucial to improving human-AI interactions and collaborations. Therefore, this dissertation aims to comprehend AI literacy and its ramifications for individuals and organizations by asking three overarching research questions. First, it asks how AI literacy can be conceptualized, measured, and enabled, laying the foundation for further AI literacy research. Second, it ventures into the effects of AI literacy on individual humans, specifically, asking how it affects their AI-related cognition (e.g., AI-related intentions and attitudes) and behavior (e.g., AI delegation behavior) in human-AI collaborations and interactions. Third, it examines the effects of AI literacy on organizations, specifically probing how the AI literacy of top management teams (TMTs) affects their organizations' AI strategy and implementation. To answer these three research questions, this thesis draws on five research articles. The first article contributes to answering the first research question by focusing on conceptualizing AI literacy for users. It conducts a systematic literature review to synthesize existing knowledge and develops a conceptual framework for AI literacy. Through this framework, the article identifies six subject areas and five proficiency dimensions of AI literacy, providing insights into users’ required literacy for the purposeful, efficient, and ethical usage of AI technologies. Moreover, it identifies and structures the existing research regarding the different learning methods to acquire AI literacy as well as the effects of AI literacy that have been discovered so far. Continuing with addressing the first research question, the second article aims to develop a measurement instrument for assessing individuals' AI literacy. Following established scale development procedures, it conducts a systematic literature review, expert interviews, and a validation study to create a measurement model containing five dimensions and 13 items. The study provides empirical support for the proposed measurement model and validates the instrument for assessing individuals' AI literacy levels. The third article addresses both the first and the second research questions. Drawing on a design science research approach, it designs an AI literacy learning experience and evaluates its effects on human cognition. The developed learning experience, a mobile learning application, “AiLingo,” targets non-expert adults to help them enhance their AI literacy. As such, it provides an enablement tool for AI literacy, completing the answer to the first research question. Moreover, the study evaluates the learning application through a between-subjects experiment. The results show that the learning experience leads to greater AI literacy advancement than a control learning experience, validating the ability to enable AI literacy efficiently, as well as that AI literacy positively influences AI-related intentions and attitudes, which addresses the second research question. From a scientific point of view, the developed design artifact (i.e., the mobile learning application) can also be viewed as a manipulation instrument, which future studies can utilize. The fourth article also addresses the second research question and focuses on how AI literacy affects human behavior in human-AI collaborations, particularly focusing on delegation decisions. It shows through a between-subjects experiment in the image classification context that AI literacy training improves humans' delegation decisions, leading to better task performance. The findings have implications for design guidelines of human-AI collaboration, emphasizing the role that potential educational features about AI functioning and human biases could have. Last, the fifth article addresses the third research question, exploring how the AI literacy of TMTs (TMT AI literacy) influences strategic AI orientation and implementation ability of organizations. Drawing on upper echelons theory, it analyzes observational data from executives' LinkedIn profiles and firm data from 10-k statements to demonstrate that TMT AI literacy positively impacts a firm's AI orientation and implementation ability. Moreover, it uncovers a moderating effect of the type of the respective firm (startup vs. incumbent) on the effect of TMT AI literacy on AI implementation ability. These five articles collectively contribute to a comprehensive understanding of AI literacy. By conceptualizing, measuring, and enabling AI literacy, as well as exploring the effects of AI literacy for individuals and organizations, they provide valuable insights into fostering effective and responsible engagement with AI technologies in diverse contexts. From enhancing individual competencies to influencing organizational strategies, AI literacy emerges as a pivotal factor in navigating the complexities of human-AI collaborations and maximizing the value of AI technologies for humans. |
||||
Alternative Abstract: |
|
||||
Status: | Publisher's Version | ||||
URN: | urn:nbn:de:tuda-tuprints-280924 | ||||
Classification DDC: | 300 Social sciences > 330 Economics 600 Technology, medicine, applied sciences > 650 Management |
||||
Divisions: | 01 Department of Law and Economics > Betriebswirtschaftliche Fachgebiete > Fachgebiet Information Systems & E-Services | ||||
Date Deposited: | 24 Sep 2024 12:07 | ||||
Last Modified: | 26 Sep 2024 07:47 | ||||
URI: | https://tuprints.ulb.tu-darmstadt.de/id/eprint/28092 | ||||
PPN: | 521684773 | ||||
Export: |
View Item |