Topic: Generative AI in Education: Adapting Training and Assessment

Lecturer: Dr. Eric Olivier

Time: Friday, May 16, 2025, 9:00 - 12:00, UTC + 8

Venue: Room 510, Xuehai Building, North Area of Nanhu Campus

Biography: Dr. Eric Olivier is an instructional designer at Aix-Marseille University (amU), where he has been actively engaged in supporting educational innovation and the integration of digital technologies since 2012. He holds a Ph.D. in mathematics and an HDR (Habilitation à Diriger des Recherches) with a specialization in the fractal geometry of measures and sets - fields in which he has published extensively and collaborated internationally. Before transitioning into educational design, he conducted postdoctoral research in France and at the Chinese University of Hong Kong, focusing on ergodic theory and chaotic dynamical systems. His academic rigor and scientific curiosity now feed directly into his pedagogical work, especially in the development of digital tools and strategies for higher education. At amU, Dr. Eric Olivier has led several large-scale initiatives, such as the “Placement Tests and Remediation” project for undergraduate science students, and the “Practicéa” program, which supports thousands of students each year in navigating amU’s digital learning environment. His work has consistently blended technological innovation with thoughtful pedagogical design, including the implementation of remote tutoring models and adaptive learning resources. He has overseen the deployment and training for plagiarism detection tools, manages the €200,000 annual Pedagogical Investment Fund (FIP), and acts as a key advisor on emerging technologies in education. With the recent rise of generative AI, Dr. Eric Olivier has been entrusted with the university’s technology watch on AI in education. He co-organized amU’s first working group on AI in teaching, and now leads institutional efforts to train faculty and guide responsible adoption of these tools.

Abstract: The implementation of training programs on the use of generative AI for faculty members at Aix-Marseille University (amU) raises questions that match the scale of this ongoing revolution. The speed at which students have embraced these new tools is the entry point to this issue. This rapid adoption presents two major challenges: on the one hand, the risk that students may develop poor study habits or even use generative AI to cheat; on the other hand, a gap with faculty members, who take longer to familiarize themselves with these tools. This creates an unprecedented situation in higher education: students’ study conditions are radically different from those experienced by their professors when they were students themselves. This is why the initial phase of AI training for AMU faculty has focused on upskilling them so they can better understand the strategies their students are likely to adopt. This first step aims to prepare for an evolution in teaching and assessment methods that must take this new ecosystem into account. The goal of this first phase of training is to enhance the skills of instructional designers and faculty members in preparation for a second phase, which will focus on setting up a sustainable generative AI infrastructure within the university. This infrastructure must support high-quality education that necessarily integrates generative AI, but without making it systematic. The balance between the use of these tools in training and assessment methods (both written and oral), from which AI could largely be excluded, is a major area of reflection. Finally, these infrastructures must meet a dual requirement: ensuring sufficient performance for academic work from undergraduate to research level while adhering to energy efficiency criteria and protecting sensitive and personal data.

Rewritten by: Mei Mengqi

Edited by: Liang Muwei, Li Tiantian

Source: School of Computer Science and Artificial Intelligence