Librarians can model best practices in engaging with AI, and support users by continually updating and communicating their own information seeking skills, particularly by modelling effective prompt engineering strategies. Users’ search strategies that worked previously, such as Boolean search terms, appear to be losing relevance as the shift towards AI in search engines prioritises different methods, primarily prompt engineering. Mastering this technical skill requires an informed, literate approach.

This peer-reviewed chapter provides a succinct survey of AI-related literature, linking traditional information literacy with its emerging form, AI literacy. It offers a brief history and definition of AI, identifies current models, and discusses drawbacks, such as hallucinations and bias. The authors emphasise informed AI use to improve functionality, while advocating for productive, ethical and safe interactions. Specificity in prompt engineering is recommended, with examples like instructive and contextual prompts. This work is targeted at an academic audience familiar with technological terminology, briefly taking a multidisciplinary approach before focusing on digital humanities applications.

Aguiar and Araújo are affiliated with a highly ranked Portuguese university. The chapter’s reference list offers great value, including widely-cited, high-quality sources that ensure the source’s accuracy and relevance. These references focus on societal risks and implications, many works of which have previously been cited by thousands of other publications, further solidifying the source's credibility.

This scholarly conference paper provides a focused approach to identifying best practices for prompt engineering, including a walkthrough of platform-specific features and pitfalls across various models. The article is tailored to an academic audience, incorporating domain-specific language to frame its strategic guidance on natural language prompts that result in quality AI responses. This source effectively weighs existing guides, while providing clear instructions for crafting advanced prompts. Readers benefit from learning to optimise results while avoiding potential shortfalls. Recommendations include advanced search techniques like batch prompting, reverse engineering, and prompt chaining.

This source represents a cross-national funding partnership between the Spanish and Luxembourgish governments. This influential article is in the 98th percentile of the Scopus citation metrics. Mid-career in an emerging field, these authors have strong h-indexes of 38 and 18, with several other related articles either recently published or set for 2025. Both are affiliated with Barcelona-based universities.

The globally-renowned public intellectual Harari explores the human history of information, tracing AI’s evolution in this newly published work. He provides accessible definitions of complex terms like misinformation (“an honest mistake”) and disinformation (“a deliberate lie”) which are sure to influence the public discourse (2024, p. 14). He posits that information is fundamental to human society, and that literacy skills imparted by trusted institutions are crucial for upholding democracies. He cites AI’s enormous positive potential in various areas, while critiquing dangers like biased decision-making by AI bureaucracies. He recommends guardrails, including literacy and regulation, to mitigate the societal consequences of unethical AI use.

Harari communicates effectively to a broad audience, writing bestsellers translated into dozens of languages. This mainstream book is valuable for information professionals as it frames the public’s understanding of AI, equipping them to respond in an informed manner. Despite criticism from academics for some viewpoints, Harari offers a more balanced perspective than Kurzweil and Suleyman, and his citation-heavy work connects with a huge audience.

This highly in-depth governmental report gives an overview of AI history, and identifies UK-based social implications of AI applications. Topics include opportunities, risks, misuse, biases, GDPR, and responsible implementation. The report acknowledges that the public are already at risk of misinformation, recommending widespread digital literacy initiatives to ensure that the public are well-versed in this emerging technology. The report takes a humanist approach, highlighting the criticality of collaboration across civil society organisations to improve AI literacy.

The House of Lords is an ancient UK parliamentary institution, with fluctuating public approval levels, similar to other governmental facets. Its appointed membership is tasked with acting on behalf of UK citizens to question public policy, producing reports through select committees. This report's credibility is bolstered by building upon various governmental and academic reports and accompanying collated oral evidence volume from expert witnesses. This report provides comprehensive appendixes, 645 high-value cited references, and the most definitions.

Kurzweil’s latest book is part of a series on the rise of AI, beginning in the 1990s. He advocates for widespread AI adoption, asserting that its positive potential is unparalleled in human history. He links social advancements like poverty reduction and healthcare improvements to information science progress. He advises that the public approach AI with caution but not resistance, emphasising digital literacy and the harnessing of AI for advancing equitable knowledge distribution. He warns that a lack of AI literacy is detrimental to societal development.

Compared to Harari and Suleyman, Kurzweil is more of an AI apologist, taking an eager role in advancing this technology. His commercial affiliations and entrepreneurial investments bias him in favour of promoting AI’s grand achievements over societal dangers. Though Kurzweil strikes some balance by critiquing the technology’s limitations, this is addressed only briefly in one late chapter, titled “Peril.” Ultimately, this citation-rich book’s value lies in providing terminology and insights into the radical pro-AI viewpoint.

This peer-reviewed article provides important context for information professionals seeking to understand how AI literacy impacts public engagement and interest in its ethical and regulatory aspects. It explores the value of information literacy from various angles, including theory, practice and policy. The social and political implications of AI are discussed, alongside the importance of regulation initiatives. The authors offer a theoretical framework titled O-S-R-O-R; Orientation-Stimulus-Reasoning-Orientation-Response. This framework links together citizen involvement, AI policy support, and information literacy.

The authors’ disclosure statement identifies that they have no conflicts of interests. Their scholastic credibility is verified through their professorial university affiliations. This resource is highly valuable in identifying the implications of AI literacy in the public sphere, in particular for readers with a strong existing knowledge base. This source illuminates the information enquiry by offering clear evidence of exactly how and why public information literacy is so critical.

This article asserts that information professionals require high AI literacy, with a focus on creativity, adaptation and interdisciplinary knowledge. The author argues that effective prompt engineering is an essential skill for library workers, a role that is rapidly evolving to incorporate AI technical competencies. He outlines several key principles for effective prompt engineering. Information professionals are advised to take an active role in supporting patron interactions with AI, in line with the CLEAR Framework.

This peer-reviewed journal article acts as an ideal source for finding similar results, due to its strong focus on recommending library staff master prompt engineering. Lo acts as Dean of the College of University Libraries and Learning Sciences at the flagship state institution, the University of New Mexico. An authority on improving AI literacy in library services, he has published many other peer-reviewed articles with a similar focus. This article takes a more accessible tone than Shiri or Register, and could be utilised by those with varying levels of AI literacy.

https://dialog.proquest.com/professional/docview/3081478187/192AFFDA2A6229DA7E0/1 (Accessed: 26 November 2024).

This dissertation offers a comprehensive look at the information science field through the lens of artificial intelligence, employing case studies to explore how low literacy impacts AI usage. The author outlines AI biases and other related harms, advocates for robust regulation, and recommends greater digital literacy for information professionals. Various ethical angles are discussed, including the impacts of algorithmic harm on minorities. The author identifies the value of harm mitigation through integrated ethics, demonstrating how to improve existing AI codes of practice, recommending a more humanist approach to AI interactions.

This paper is aimed at academics, submitted as part of a Doctorate of Philosophy for the Information School at the reputable institution, University of Washington. The reading committee are established academics. This source appears well-situated within the existing body of research, as it has shared references with 3101 other documents, listing 534 references. Compared to Harari or Lo, this source is less accessible for those new to this topic.

This recently published, peer-reviewed article outlines AI literacy concepts with an academic audience in mind, providing tools to better communicate about AI. This source enriched the search by defining terms, highlighting the importance of constructing a cross-disciplinary AI basic vocabulary. This high-level work delves deeper into existing AI literacy frameworks than Clarisó and Cabot’s article, ultimately producing a more comprehensive proposed taxonomy that is clearly presented in table format. The article also identifies many more vocabulary gaps, aiming for a conceptual framework for metaliteracy across various user communities, including library patrons. Information professionals who take an informed stance, understanding AI from a conceptual standpoint, are best positioned to mitigate user risks.

The peer-reviewed hosting journal Digital Library Perspectives is linked to the acclaimed, high-impact global nonprofit library organisation OCLC. 15 out of the 17 cited references have been peer-reviewed, lending further credibility to this champion source.

The primary author, Suleyman, has had a huge impact on the field of AI, founding both DeepMind and Inflection AI. This work, aimed at a general audience, charts the rapid acceleration of the technology he has helped develop. While he posits that humanity can positively leverage AI capabilities, this book carries dystopian warnings. He cautions that AI can exploit social divisions, urging democratic societies to stay ahead in the AI arms race, painting a sometimes bleak and terrifying future. He claims that AI will replace “intellectual manual labour,” cataclysmic to the future role of information professionals (2023, p. 31).

Despite his clear commercial affiliations, he recommends significant limitations on its use. He advises regulation, transparency, containment and information literacy. He takes a much more cautious, borderline nihilistic approach than Kurzweil, and a less historical approach than Harari. This balanced work provides a multitude of well-supported talking points useful for librarians communicating AI literacy to patrons, alongside many citations.

Azamfirei, R., Kudchadkar, S. R., and Fackler, J. (2023) ‘Large language models and the perils of their hallucinations’, Critical Care, 27(120), p. 1. Available at: https://doi.org/10.1186/s13054-023-04393-x

Chiang, C.-W. and Yin, M. (2022) ‘Exploring the Effects of Machine Learning Literacy Interventions on Laypeople’s Reliance on Machine Learning Models’, Proceedings of the 27th International Conference on Intelligent User Interfaces (IUI). Helsinki (Finland), 22-25 March. New York: Association for Computing Machinery, pp. 148-161. Available at: https://doi.org/10.1145/3490099.3511121

Fischer, J. E. (2023) ‘Generative AI Considered Harmful’, Proceedings of the 5th International Conference on Conversational User Interfaces (CUI 2023). Eindhoven (Netherlands), 19-21 July. New York: Association for Computing Machinery, pp. 1-5.

Available at: https://doi.org/10.1145/3571884.3603756

Lin, Z. (2024) ‘How to Write Effective Prompts for Large Language Models’, Nature Human Behaviour, 8, pp. 611-615. Available at: https://doi.org/10.1038/s41562-024-01847-2

Zamfirescu-Pereira J., Wong R., Hartmann B., Yang Q. (2023) ’Why Johnny can’t prompt: how non-AI experts try (and fail) to design LLM prompts’, Proceedings of the 2023 CHI conference on human factors in computing systems (CHI ’23). Hamburg (Germany), 23-28 April. New York: Association for Computing Machinery, pp. 1-21. Available at: https://doi.org/10.1145/3544548.3581388