MinIAtures: Turning AI Guidelines into Daily Habits Through Student-Led Peer Coaching

Team name: Beyond the bot

Use of AI tools:

During this project, our team used the following AI tools : Gemini, Claude, NotebookLM, and ChatGPT. These tools were used for:

-refining and clarifying conceptual ideas;

-generating practical examples for our microlearning videos to demonstrate the correct use of AI to first-year students and serve as tutorials.

These tools supported our work by helping us visualize the practical application of our theoretical framework. However, all final ideas, pedagogical choices, and reflections were critically reviewed and validated by our team.

External feedback & contributions :

Several people and discussions contributed to the improvement of our project:

Caroline Beslin (one of the keynote speakers): Gave our team valuable insights on improving the structure of our project , which helped us improve our approach to motivating first year students and cultivating/preserving learner’s autonomy.

Dr.Christelle Scharff : Advised us to be more precise with our contribution so the reader can envision the project as intended.

Bader daraa: Provided critical feedback regarding the generation of mock exams, prompting us to clarify our data sources and evaluation methods.

Hans Zúñiga: Validated our core concept, confirming that a guided learning process is a viable alternative to the excessive use of AI in academic settings.

Nipun ranchohd Navadia : Asked valuable questions about sustaining students engagement and methods of training as well as evaluation.

Frederic Ooms, Dominique Verpoorten and the Digital Expérience team: Provided the institutional framework, co-creation structure, and pedagogical alignment with the university's goals.

Initial contribution: Please provide the link (URL) to your initial contribution on the platform: MinIAtures: Turning AI Guidelines into Daily Habits Through Student-Led Peer Coaching

Final contribution:

Title:

MinIAtures: Turning AI Guidelines into Daily Habits Through Student-Led Peer Coaching

Introduction

The integration of generative Artificial Intelligence in higher education has created a significant challenge when it comes to academic integrity and cognitive depth. Our contribution addresses the core dilemma of : does the tool train the student, or does the student train the tool?. To resolve this, we are implementing 'MinIAtures', a multimedia microlearning series where AI functions as a study coach for first-year students. Designed by students and supported by faculty, this initiative aims to transform abstract institutional guidelines into actionable, daily habits.

Context and challenges

The blind use of AI, often referred to as "cognitive offloading," threatens the development of critical thinking. While institutions provide high-level rules, such as the ULiège Charter, students often struggle to apply these principles practically. There is a critical need to transition students from passive consumers of AI-generated content to "vigilant editors" who can critically refine machine outputs.

The Constrained AI Tutor Model

Instead of allowing AI to function as a tool that delivers direct answers, our proposal introduces a "constrained AI tutor" model. This approach intentionally generates "pedagogical friction," ensuring that the AI guides the user's reasoning process and keeps the student cognitively engaged.

Through a series of short, practical modules hosted on the university's learning platform, students are trained in:

Applying and experimenting with advanced prompting using the Socratic method and the RCOC framework.

Course organization, synthesis, and the generation of structured flashcards.

Practicing with mock exams and language conversation.

Crucially, this is delivered through a student-led peer-coaching program, scaling high-quality, responsible support to all students.

Reflection on the process

During Phase 2, our contribution evolved significantly through the feedback and reflections shared during discussions and keynotes. Initially, our proposal focused heavily on the theoretical application of the Socratic method. However, feedback from the many commenters highlighted a practical concern: how mock exams, practice flashcards and summaries would be generated and what data they would be based on. This prompted us to refine our methodology, clarifying that students must input their own course materials into AI tools (such as utilizing their own old notes or course material that professors consent to share to be exploited in such way) to ensure data privacy and academic relevance.

Furthermore, our session with Miss Caroline Beslin fundamentally shifted our perspective on project communication. We realized that to ensure long-term adoption, we needed to emphasize the "autonomy" and « experience » students gain over their learning process by following the capsules made by students who have already passed those subjects. Finally, other feedback reinforced that explicitly programming AI to guide the learning process, rather than provide immediate solutions, is the most viable path to controlling excessive AI dependency. This process greatly strengthened our final proposal, making it more realistic, secure, and deeply connected to the lived experiences of first-year university students.

Conclusion

By combining multimedia microlearning with a human-centered peer-coaching framework, MinIAtures bridges the gap between institutional charters and daily academic practices. Our project ensures that the learning process remains indisputably human, fostering active learning while safely integrating AI as a reasoning partner.

Video pitch

Please provide a link to your 1-minute video pitch presenting your work: this is a sneak peak into one of our microcapsules : https://drive.google.com/file/d/1pAcpS7jUmSfycvSw8biykIBEHtNw_hGY/view