The pursuit of sustainability is not just a trend or an organisational strategy—it’s a responsibility that we all share. But how can we effectively incorporate sustainability into our teaching materials and empower our students to become agents of change?
Imagine teaching where AI is not just a concept but a practical tool for integrating sustainability into our curriculum.
AI presents an exciting opportunity in education (see Morrison, 2022). Imagine teaching where AI is not just a concept but a practical tool for integrating sustainability into our curriculum. In this post, we’ll explore how AI can help the way we teach and promote sustainability within our universities with examples from Metropolia’s prototyping with an SDG Moodle plugin. This post is part of a running teacher blog series on sustainability in education.
A Journey Through AI’s Evolution in Education
What the current fast development of AI means in the context of education is that we no longer need to develop AI tools for every particular need such as essay grading, assignment planning or generation of presentations. Large language models (LLMs) are mature enough to be used in a variety of different tasks as they are. The best part is that LLMs might be clumsy today, but they are being improved every day.
The best part is that Large language models might be clumsy today, but they are being improved every day.
Let’s take a moment to reflect on the evolution of AI within education. From rule-based approaches (see Uibo et al., 2017) to sophisticated neural networks (see Ndukwe et al., 2020), AI has transformed the way we teach and learn. But what does this mean for sustainability education, and how can AI support us as educators?
We’re witnessing a shift—a shift towards AI becoming a valuable ally in our quest to integrate sustainability into our teaching materials. By leveraging AI, we can create engaging, interactive content that fosters a deeper understanding of sustainability concepts among our students. In terms of AI research, there have been plenty of recent developments both in automatic creation of high-quality content (see Xu et al., 2020; Koppatz et al., 2022) and in automated analysis of sustainability goals (see Kharlashkin et al., 2024).
The shift in AI becoming more and more useful in a variety of different fields can be explained by the amazing generalisation capability of the current state-of-the-art neural network architecture, namely the Transformer architecture (Vaswani et al., 2017). This capacity of generalising has led to the emergence of intelligence in LLMs (see Anil et al., 2023). In other words, we have entered an era where AI models possess capabilities that have not been explicitly designed by their developers. Their capabilities emerge from the combination of the sheer volume of data they have been trained on and their size. The bigger the LLM, the more it seems to be able to learn from data.
Moodle Plugin: Empowering Teachers for Change
Sustainability isn’t just a subject—it’s a mindset. As educators, we have the power to instill values of sustainability in our students and inspire them to make a positive impact on the world. By integrating sustainability into our teaching materials, we can empower our students to become active participants in building a more sustainable future. But we also recognize the challenges. It can be daunting to know where to start when it comes to incorporating sustainability into our curriculum. That’s where AI comes in as a tool that not only assists in creating engaging teaching materials but also ensures that sustainability concepts are seamlessly integrated into every lesson.
AI comes in as a tool that not only assists in creating engaging teaching materials but also ensures that sustainability concepts are seamlessly integrated into every lesson.
The way the Moodle plugin works is that it uses an LLM to analyse teachers’ slides. Based on this analysis, the plugin recommends suitable sustainable development goals for each lecture and gives ideas on how to incorporate each goal with the teaching material. This is done by automatically recommending talking points and assignments for each SDG that also relate to the teacher’s own material. When a user starts using the plugin, they will see the view depicted in Figure 1.
Figure 1: SDG analysis
As seen in Figure 1, the tool can create a summary of the content of slides and recommend SDGs that are suitable for the class in question. The user can pick their desired SDG and get some information on how the SDG relates to the slides and what kind of an assignment the teacher could give their students to practise both what has been taught in class and the desired SDG. This can be seen in Figure 2.
Figure 2: Further analysis for SDG 12
The Moodle plugin also makes it possible to have a free form chat with the LLM as seen in Figure 3. In this case, the LLM knows the contents of the slides and the teacher does not need to reupload them as they would need to do if they were using ChatGPT.
Figure 3: Chat feature
These functionalities make it easy for teachers to incorporate sustainability into their teaching. The chat feature makes it easy for teachers to try out things that aren’t directly supported by the interface without ever needing to leave Moodle.
New Features are Planned in Workshops
Embracing AI as a tool for change, educators have the potential to inspire a new generation of sustainability champions. So far, the development of the Moodle plugin has been in the hands of the development team. However, we are interested in making the plugin as useful as possible. For this reason, we are organising workshops for our teachers in Metropolia with the objective of gathering as much user experience feedback as possible. This way we can improve the overall usability of the plugin and add the kind of features that our teachers would like to see in Moodle.
Our journey with AI in education is just beginning, and the possibilities are vast.
Our journey with AI in education is just beginning, and the possibilities are vast. By leveraging AI’s capabilities, we can further develop traditional teaching methods and inspire our teachers and students alike to become active agents of the gospel of sustainability. Let us seize this opportunity to transform education and, in doing so, shape a world where sustainability is at the core of every decision. Together, we can inspire, educate, and empower our students to build a more sustainable future.
Writers
Tricia Cleland Silva serves as the coordinator for the Metropolia Sustainability in Education team. She is a senior lecturer in the master degrees of Health Business Management and Leadership and Development in Nursing. She holds a PhD in Management and Organization. Her co-created method of Collaborative Story Craft and Story Mediation inspires her roles in sustainable development and inclusion within higher education and various communities of practice. She is also a parent, owns and runs a family business with her partner and an immigrant professional from Canada.
Mikä Hämäläinen works as an AI project manager in the Strategy and Development Services of Metropolia University of Applied Sciences. His PhD focused on creative natural language generation, and he has published numerous research papers in a variety of fields such as natural language processing, computational creativity and digital humanities.
Lev Kharlashkin is one of the developers of the AI Moodle plugin at Metropolia University of Applied Sciences, where he is currently pursuing his bachelor’s degree in Smart IoT Systems. He is interested in integrating intelligent systems into educational platforms. Outside of his academic and professional pursuits, Lev enjoys exploring new technologies and contributing to open-source projects. He has a keen interest in the future of smart systems and their potential impact on the world.
Melany Macias is one of the developers of the AI Moodle plugin, and she works for the Strategy and Development Services department. Melany is currently a 4th-year IT Engineering student at Metropolia, with her thesis centred on the development of this plugin and as she is near her graduation, she also plans to pursue a master’s degree in Artificial Intelligence. Beyond her academic commitments, she actively participates in social gatherings, hackathons and startup events, showcasing her dedication to applying technology in practical ways.
References
Anil, R., Dai, A. M., Firat, O., Johnson, M., Lepikhin, D., Passos, A., & Wu, Y. (2023). Palm 2 technical report. arXiv preprint arXiv:2305.10403.
Kharlashkin, L., Macias, M., Huovinen, L., & Hämäläinen, M. (2024). Predicting Sustainable Development Goals Using Course Descriptions from LLMs to Conventional Foundation Models. Journal of Data Mining & Digital Humanities.
Koppatz, M., Alnajjar, K., Hämäläinen, M., & Poibeau, T. (2022). Automatic Generation of Factual News Headlines in Finnish. In Proceedings of the 15th International Conference on Natural Language Generation (pp. 100-109).
Morrison, R. (2022). Large Language Models and Text Generators: An Overview for Educators. ERIC Institute of Education Services.
Ndukwe, I. G., Amadi, C. E., Nkomo, L. M., & Daniel, B. K. (2020). Automatic grading system using sentence-BERT network. In Artificial Intelligence in Education: 21st International Conference, AIED 2020, Part II 21 (pp. 224-227). Springer International Publishing.
Uibo, H., Rueter, J., & Iva, S. (2017). Building and using language resources and infrastructure to develop e-learning programs for a minority language. In Proceedings of the joint workshop on NLP for Computer Assisted Language Learning and NLP for Language Acquisition (pp. 61-67).
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., & Polosukhin, I. (2017). Attention is all you need. Advances in neural information processing systems, 30.
Xu, P., Patwary, M., Shoeybi, M., Puri, R., Fung, P., Anandkumar, A., & Catanzaro, B. (2020). MEGATRON-CNTRL: Controllable Story Generation with External Knowledge Using Large-Scale Language Models. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) (pp. 2831-2845).
Ei kommentteja