The Leprechaun and the Librarian (AI can grant wishes)

A presentation comparing interactions with AI to speaking with a librarian through the analogy of AI as a leprechaun who "gives you the gold".

1. The Leprechaun and The Librarian

2. You are not an anti-tech crowd. AI is not simply another tool.

3. You are not an anti-tech crowd. AI is not simply another tool. Human Intelligence Experience → Understanding → Meaning LLM Intelligence Data → Pattern → Prediction

4. Careful what you ask for…. source: https://www.reddit.com/r/ChatGPT/comments/11qxlyr/chatgpt_calculates_the_last_10_digits_of_pi/

5. Shhh! What is a librarian? (Beyond the bun.) Rich history of libraries for centuries Training in information theory and information-seeking behavior Training in classification, storage & retrieval, acquisition selection policies, information freedom, intellectual property Data management, archival, preservation Embedded librarian- a person not a place!

6. Print is no longer KING! Formats: digital and physical (metadata, physical resources, hardware, software, XR and AI evolving with IT)

7. AI is like a leprechaun who grants what you wish for!

8. What is the difference between the librarian and the leprechaun when we ask for information?

9. Librarian / Library AI Leprechaun / Internet The library is a living organism. LLM training corpus grows a monster machine.

10. Librarian / Library AI Leprechaun / Internet The library is a physical community gathering space. The Internet is a virtual web.

11. Librarian / Library AI Leprechaun / Internet The librarian strives to lead toward critical thinking. The leprechaun strives to give you the gold or “fool’s gold”.

12. Librarian / Library AI Leprechaun / Internet Prioritizes accuracy, authority, ethics Prioritizes engagement and corporate success/profits

13. Librarian / Library AI Leprechaun / Internet Classification is consistent, planned and purposeful Classification is random, through statistical prediction

14. Librarian / Library AI Leprechaun / Internet Acquires & Organizes meaningful information Predicts meaning word by word through algorithms

15. Librarian / Library AI Leprechaun / Internet Transparent Access (Free access to the community) Opaque/Paywall Access

16. Librarian / Library AI Leprechaun / Internet Preservation & “weeding the garden” Ongoing information hygiene Invisible “chaos” & Dead Internet Theory

17. Librarian / Library AI Leprechaun / Internet Verifies sources for authority/accuracy/ Aesthetics Values history Cannot verify- more AI content now generated than human content Recreates history “artificial”

18. WAIT! What is information anyway? Information is not equal in meaning or priority. Information is not free. Information is not always accurate. Information is contextual and changing.

19. 1. Books Are For Use 2. Every Reader his/her Book 3. Every Book Its Reader 4. Save The Time Of The Reader 5. The Library Is A Growing Organism Five Laws of Library Science

20. Do no harm/obey 1. A robot may not injure a human being or, through inaction, allow a human being to come to harm. 2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law. 3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law. Asimov’s Laws of Robotics/Rules of Corporate AI Profit/Win the AI Race (huge alignment issue) 1. Maximize algorithm for profit 2. Minimize operating costs 3. Please the user (regardless of information quality)

21. Talking to the Leprechaun

22. Anatomy of a Prompt Context/Role Who the AI should be or w hat perspective to take. Task What you want the AI to do. Constraints Specify format, length, tone, or limitations

23. Example Ambiguous Prompt "Tell me about AI in libraries" Clearer Prompt " [Context] You are a library science professor. [Task] Explain three ways AI tools differ from traditional library databases when students are doing research. [Constraints] Keep it under 150 words and use simple language suitable for undergraduates. [Example] For instance, discuss aspects like source verification, search methodology, or result consistency."

24. Final Thoughts on Prompting Be specific Ask it to show its work Ask for multiple perspectives Fact-check everything Iterate and refine Experiment!

25. We expect more from technology and less from each other. We create technology to provide the illusion of companionship without the demands of friendship. ~Sherry Turkle

26. References Amaral, M. (2024, January 30). AI is creating fake historical photos. Substack. https://marinaamaral.substack.com/p/ai-is-creating-fake-historical-photos Anthropic. (2025). Claude. https://www.claude.com/product/overview big_chestnut. (2023). ChatGPT calculates the last 10 digits of Pi [Reddit post]. r/ChatGPT. https://www.reddit.com/r/ChatGPT/comments/11qxlyr/chatgpt_calculates_the_last_10_digits_of_pi/ Bloomberg News. (2025, September 29). AI data centers are sending power bills soaring. https://www.bloomberg.com/graphics/2025-ai-data-centers-electricity-prices/ Center for Humane Technology. (2025). Center for Humane Technology. https://www.humanetech.com/ Center for Humane Technology. (2025). Your Undivided Attention [Podcast]. https://www.humanetech.com/podcast DataCamp. (2024, July 23). A beginner’s guide to ChatGPT prompt engineering. https://www.datacamp.com/tutorial/a-beginners-guide-to-chatgpt-prompt-engineering DGTrainingLand. (2025, September 3). Complete guide to prompt engineering for beginners. Medium. https://medium.com/%40dgtrainingland/complete-guide-to-prompt-engineering-for-beginners-e634ae59a5be Emerald. (2023, July 12). So you want to be a sorcerer in the age of mythic powers… (the AI episode) [Audio podcast episode]. In The Emerald. Spotify. https://open.spotify.com/episode/22QF1duMlwvws0QbMFJVLA

27. References Getty Images. (1953, August 8). British archaeologist Lankester Harding in the caves on the north-west shores of the Dead Sea, Jordan [Photograph]. Picture Post. https://www.gettyimages.com/detail/news-photo/british-archaeologist-lankester-harding-in-the-caves-on-the-news-photo/2663881 Google. (2025). Gemini. https://gemini.google.com/ International Energy Agency. (2025, April 10). Energy and AI. https://www.iea.org/reports/energy-and-ai Learn Prompting. (2024, October 23). Prompt engineering guide: The ultimate guide to generative AI. https://learnprompting.org/docs/introduction MIT Sloan EdTech. (2025). Effective prompts for AI: The essentials. https://mitsloanedtech.mit.edu/ai/basics/effective-prompts/ Midjourney. (2025). Midjourney. https://www.midjourney.com/ Nosta, J. (2025, July 15). AI and the architecture of anti-intelligence. Psychology Today. https://www.psychologytoday.com/us/blog/the-digital-self/202507/ai-and-the-architecture-of-anti-intelligence/amp Nosta, J. (2025, October 16). Artificial intelligence and the inversion of intelligence. Psychology Today. https://www.psychologytoday.com/us/blog/the-digital-self/202510/artificial-intelligence-and-the-inversion-of-intelligence/amp

28. References OpenAI. (2025). ChatGPT. https://chatgpt.com/ Postman, N. (1992). Technopoly: The surrender of culture to technology. Knopf. Ranganathan, S. R. (1931). The five laws of library science. Edward Goldston, Ltd. Rubin, R. E. (2015). Foundations of library and information science (4th ed.). ALA Neal-Schuman. Sivarajkumar, S., Balakrishnan, R., Chen, L., & Wu, S. (2024). An empirical evaluation of prompting strategies for large language models in clinical information extraction. JMIR Medical Informatics, 12 (1), e55318. https://medinform.jmir.org/2024/1/e55318 Trad, F., Yammine, R., Charafeddine, J., Chakhtoura, M., Rahme, M., El-Hajj Fuleihan, G., & Chehab, A. (2025). Streamlining systematic reviews with large language models using prompt engineering and retrieval augmented generation. BMC Medical Research Methodology, 25 , 130. https://bmcmedresmethodol.biomedcentral.com/articles/10.1186/s12874-025-02583-5 Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E., Le, Q., & Zhou, D. (2022). Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems (NeurIPS 2022). https://arxiv.org/abs/2201.11903 Wilkins, J. (2025, September 28). First responders are being overwhelmed by data center fires. Futurism. https://futurism.com/future-society/fire-fighters-data-centers

29. The Leprechaun and The Librarian Rose Hill / roropillow Valerie Hill / Valibrarian