黑料福利网

Joris Remmers, associate professor at Mechanical Engineering and AI learning connector at ALT, opened the session by presenting how AI in education is currently discussed in Dutch media. The examples illustrated a broad public debate, ranging from concerns about the impact of AI on teaching to questions about whether AI offers meaningful efficiency gains or introduces new risks.

Hildo Bijl (assistant professor at Mathematics & Computer Science) presented the , which aims to reduce students鈥� reliance on AI through improved course design. Based on self-reported data from students using the , AI usage ranged from 鈥渆qual鈥� to 鈥渁 lot less.鈥� Two-thirds of students indicated that well set-up and structured education can reduce their tendency to resort to AI when studying.

Natalia Sidorova (assistant professor at Mathematics & Computer Science) shared insights from the , involving 593 students. While not the main focus, early observations suggest there could be gender differences in AI use. A key takeaway from students is the need for clear guidelines on the acceptable use of AI. Next steps in their project focus on effective and responsible AI use, as well as strengthening critical thinking.

Ivo Filot (assistant professor at Chemical Engineering & Chemistry) focused on the 鈥攁 controlled environment in which teachers retain oversight of large language models, as part of the . According to Filot, Alexandria is currently the second-most preferred tool among their students, after ChatGPT.

His observations echoed Sidorova鈥檚 findings: students benefit from clear guidance on how AI can be used responsibly. Filot similarly emphasized that guidelines, AI literacy, and critical thinking are essential, alongside skills such as verification and prompt engineering.

As part of this, he introduced the 鈥�5 Commandments for AI in Learning鈥�:

The role of AI in assessment was explored by C膬lina Ciuhu (assistant professor at Electrical Engineering), who presented the , which monitors prompts and supports grading, feedback, and surveys. As part of this work, they also conducted a workshop with EE colleagues at their Education Day on designing course-specific AI models for assessment.

Ciuhu's colleague Mitrofan Curti (assistant professor at Electrical Engineering) had previously presented related work during the 2nd jam session in December 2025, highlighting the continued development of this line of research. That earlier session also featured a contribution from Jos Elfring (assistant professor at Mechanical Engineering), who presented a project on an .

Peter Ruijten-Dodoiu (assistant professor at Industrial Engineering & Innovation Sciences) addressed the challenge of providing meaningful feedback at scale. Their project focuses on :

Providing high-quality formative feedback to written reflections is time-intensive. And with bigger class sizes, feedback becomes delayed, shallow, or sporadic. This creates a structural gap between pedagogical intentions and actual teaching practices. We鈥檇 like to explore the possibilities of AI in our efforts to close this gap.

He also highlighted a broader opportunity for AI in education:

Most educational uses of AI focus on questions that have objectively correct answers. We believe that AI also makes it possible to rethink formative feedback on written reflections.

In the final presentation, Joris Remmers reflected on how his teaching has evolved alongside the rise of large language models. Starting in 2023 with exploring how students use tools like ChatGPT, his approach progressed to embedding AI in education in 2024, to using LLMs as tutors and a in 2025.

By 2026, this journey has led to alternative assessment methods, as their team of two teachers and six teaching assistants typically conduct around 200 oral exams. His central message: 鈥淎dapt your teaching to a world with LLMs,鈥� which requires rethinking learning outcomes and redesigning assessment strategies, since existing approaches are no longer sustainable or scalable.

The session concluded with an invitation to the 黑料福利网 AI & Education Internal Summit on 11 May 2026, where the conversation on AI in education will continue.