Generative Artificial Intelligence and Academic Integrity - LIR HEAnet User Group for Libraries
1 like501 views
Dr. Thomas Lancaster discusses the implications of generative artificial intelligence (AI) on academic integrity and library services, emphasizing the potential for AI to enhance content creation while also posing ethical challenges. Key concerns include contract cheating, where students use AI to complete work without learning, and the need for librarians to develop policies regarding AI usage and reference practices. The presentation highlights the importance of understanding the limitations and risks associated with AI in academia.
Generative Artificial Intelligence and Academic Integrity - LIR HEAnet User Group for Libraries
- 1. Generative Artificial Intelligence and Academic Integrity Dr Thomas Lancaster Imperial College London UK “We Need To Talk About AI” (7 June 2023) LIR HEAnet User Group for Libraries
- 2. Dr Thomas Lancaster Computer Scientist at Imperial College London Academic integrity researcher and speaker http://www.thomaslancaster.co.uk Twitter: @DrLancaster Machine generated image
- 3. A student pays or uses a third party to complete assessed work for them Contract Cheating
- 4. Many of the interventions the sector have put into place surrounding contract cheating are also relevant for the misuse of artificial intelligence Contract Cheating
- 5. • Generative AI is a technology that generates new, unique content like book summaries or reading suggestions • It operates using a system that improves itself by creating and evaluating its own work, trying to mimic real-world data • It has the potential to enhance library services by creating personalized content or recommendations • However, it's crucial to be aware of ethical issues, as this technology can also create deceptive or false information Generative Artificial Intelligence (not my descriptions)
- 6. “What is generative AI? Explain in short bullet points in a way that's suitable for librarians”
- 7. “Simplify this and reduce this to four main bullet points”
- 8. “Simplify further and make sure the response is tailored to librarians”
- 9. Working with technology like ChatGPT is not always easy or fool proof!
- 10. • Not “intelligent” in the human sense • Trained on large quantities of data • Follows prompts and directions • Can mimic structures and patterns Generative Artificial Intelligence
- 11. • AI does the work – students do not learn • Students do not understand the limitations of AI • Students breach academic integrity The Dangers…
- 12. Working with AI well requires a logical brain
- 13. Working with AI well requires the ability to evaluate and critique information
- 14. • Limited access to current information • Formulaic responses • Numerical answers can be incorrect • List sources that do not exist Possible AI Limitations Note – there are routes around many of these limitations
- 15. What do students say they use ChatGPT for? editing learning ideas translation time saving
- 16. Where does academic integrity come in? editing learning ideas translation time saving
- 18. What Do Librarians Need To Know? Policy – what is allowed? What to do if a student asks about a non-existent source (How) should AI be referenced?
- 19. Let’s make sure that everyone has the tools they need to work with integrity with future AI
- 20. Connect With Dr Thomas Lancaster My Email: thomas@thomaslancaster.co.uk My Website: http://thomaslancaster.co.uk My Blog: http://thomaslancaster.co.uk/blog My LinkedIn: https://uk.linkedin.com/in/thomaslancaster My Twitter: @DrLancaster Slide Archive: http://www.slideshare.net/ThomasLancaster