Contributor: Mari Cruz García Vallejo
Name of the recipe: ¡Música, maestro!: Limitations and risks of the AI
Ingredients (any equipment or supplies needed for the activity)
A free account in Suno, an institutional account in Copilot or a free account in any other AI chatbot (e.g,, ChatGPT, Claude, etc.)
Method (what you do)
This synchronous learning activity is designed for higher education lecturers who have never used generative Artificial Intelligence (GenAI) in their teaching practice and may feel uncertain or reluctant about adopting GenAI tools. It can be integrated into both in-person and online lessons.
The activity was originally delivered as part of the introductory session of the module CETD24 - Cómo entrenar a tu dragón: la inteligencia artificial generativa como herramienta para mejorar el aprendizaje en entornos online e híbridos ("How to Train Your Dragon: Generative Artificial Intelligence as a Tool to Enhance Learning in Online and Hybrid Environments"). This module aims to provide lecturers at the University of Las Palmas de Gran Canaria (ULPGC) with the foundational skills and knowledge to use GenAI to enhance learning and assessment. However, the activity can be adapted for different contexts and groups of learners.
After an introductory session on the basics of generative AI, participants are divided into groups. Each group works through learning materials that outline the main limitations, risks, and social and ethical challenges associated with GenAI. Engaging with this information is crucial, as each group is asked to create a song using the text-to-music AI program, Suno.
The song must meet the following requirements:
- Musical Style and Arrangement: Participants can choose any musical style, arrangement, or imitation of real singers for their song.
- Lyrics Content: The lyrics should summarise the main limitations, risks, and ethical challenges posed by AI, not only for education but for society as a whole.
- Character Limit: Lyrics must not exceed 3,000 characters (this limit comes from the free version of Suno)) and must include at least two stanzas and a chorus.
- Song Duration: The song must be at least 1 minute 20 seconds and no longer than 4 minutes.
Each group first drafts their lyrics, summarising the limitations and challenges of AI. They are then allowed to use AI tools, such as Copilot or any other AI chatbot, to refine and adapt the lyrics to their chosen musical style (Suno can also assist in this adaptation). Once the songs are published, participants vote for the best song.
Examples of the songs produced by the participants belonging to the S1 academic year 2024-2025 presentation of the course:
Song "La IA nos engaña" (The AI is cheating us) : https://suno.com/song/dc06ed5b-35dc-42be-b5ee-eb6db6197a18
Canción "La máquina imperfecta" (The imperfect machine): https://suno.com/song/b026c542-f230-439c-878f-4a244b2a7065
Canción "Sombras de la IA" (Shadows of the AI): https://suno.com/song/f3c2e7f1-d06a-48c7-ae48-c9b28d72669d
Special Notes:
For learners who are new to AI and are developing an understanding of AI literacy and AI competency frameworks, it is advisable that the instructor provides materials outlining the main limitations and risks associated with AI. These materials may include issues such as impacts on copyright, the need for a regulatory framework, and ethical considerations related to AI.
Ensure that participants receive clear instructions regarding the requirements for the song lyrics. This should include the specific role that AI will play—whether it is used to enhance participants' original lyrics or to generate the lyrics autonomously.
After all groups have listened to each other’s songs and voted for the most original one, the learning activity should conclude with a debate. This discussion will delve deeper into the limitations and risks of AI, as well as its broader societal impacts.
Acknowledgements
References (links to scholarly articles, conference presentations etc where people can find out more if needed)
Creative Commons License that applies - https://creativecommons.org/share-your-work/
This work is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
