GenAI pilots at the Faculty of Science of the University of Amsterdam

by Emma Wiersma, University of Amsterdam (UvA), The Netherlands.

The rise of generative Artificial Intelligence (genAI) has a significant impact on academic education. That’s why the Teaching & Learning Centre Science of the University of Amsterdam launched a project integrating genAI, specifically ChatGPT, into the teaching process. The project will run for a year and marks a major step forward in using the potential of artificial intelligence to enhance and personalise learning.

Targeting a select number of courses within the Faculty of Science, this project aims for close collaboration between teaching teams, AI experts from the faculty, and didactical experts from the TLC Science. The aim is to effectively deploy ChatGPT in teaching, enhancing student learning and enriching the educational experience.

Pilots

In total, 12 pilots will be running from January until December 2024. From these pilots, 11 pilots  target students and 1 pilot targets lecturers.

Each of the 11 student pilots has its own focus. For example, to let students use ChatGPT as a writing coach, to give feedback on research questions, to support with programming exercises, to expose students to ideologically different counter arguments during a debate, to analyse a case from an external stakeholder or to help students develop their professional skills.

The 1 lecturer pilot focuses on reducing their workload. GenAI is being used to improve the constructive alignment within their course (aligning their assessment, learning goals and learning activities) and to create presentations and other visualisations for their education. Both cases work for existing and new courses and result in immediate efficiency.

First results

We have six ways to monitor the results of each pilot:

  1. An online midterm evaluation with the students halfway throughout the course.
  2. An online midterm evaluation with the lecturers and supervisors halfway throughout the course.
  3. An online final evaluation with the students at the end of the course.
  4. An online final evaluation with the lecturers and supervisors at the end of the course.
  5. A focus group with the students at the end of the course.
  6. A focus group with the course coordinators after the course.

Currently, six pilots are finished. The first results indicate that the students appreciate the immediate feedback from the genAI tool, it makes their learning process more efficient and faster, it encourages reflection and deeper thinking, it helps them when they get stuck, and it can be useful as a jumpstart. Concerns are the correctness of the output, to lose authenticity of their own writing style, and to get lazy and too dependent on genAI.

The coordinators, lecturers and supervisors mentioned that the feedback from the genAI tool reduced their workload and that it is an added value for their course.

Microsoft Azure

One of the unique aspects of this project is that students and lecturers of the participating courses will have access to ChatGPT via Microsoft Azure. This approach not only ensures user security and privacy by using a shielded cloud environment, but also ensures seamless integration of this cutting-edge technology into teaching practices. This strategic choice adequately addresses data and privacy concerns, which is essential for applying these tools in a university setting.

More information

For more information and to keep updated about the pilots, you can contact tlc-science@uva.nl or visit this webpage: https://tlc.uva.nl/en/article/ai-pilots/. The webpage will be updated throughout the year.

Author

Emma Wiersma, Project Leader Educational Innovation & Educational Advisor Blended Learning, Teaching & Learning Centre Science, University of Amsterdam (UvA), The Netherlands