Manuel Sudau


Last Name

Sudau

First Name

Manuel

ORCID

0000-0003-3230-9160

Organisational unit

06167 - Curriculumsentwicklung / Curriculum Development

Filters

2023 - 20247
Reset filters

Search Results

Publications 1 - 7 of 7
  • Assessing the Potential of AI for Scientific Writing at ETH Zurich: Recommendations and Findings
    Item type: Educational Material
    Mihálka, Réka; Sudau, Manuel; Paschke, Melanie (2024)
  • GenAI’ze your classroom
    Item type: Educational Material
    Sudau, Manuel; Paschke, Melanie; Mihálka, Réka (2024)
    With the release of ChatGPT-3 in fall 2022 (OpenAI, 2022), technological developments in the field of artificial intelligence have posed a major challenge for the higher education landscape. Generative AI (GenAI) and the new AI-based tools such as Microsoft Copilot, GitHub Copilot, ChatGPT, Google Gemini, Wolfram Alpha and many more are here to stay. Their emergence and widespread availability are creating benefits and obstacles to teaching activities. As a lecturer you have probably asked yourself: Do I have to do something about GenAI in my course? If you must or want to do something, what can you do? During our work in the Department of Environmental Systems Science (D-USYS) and the Zurich-Basel Plant Science Center (PSC), it quickly became clear that one of the most important needs of our lecturers was to integrate the training and use of AI-based tools into the subject-specific context of existing courses. The necessary considerations included, for example, the adaptation of course descriptions in the course catalog, the definition of the use of AI in learning, performance and examination tasks, as well as the redesign of teaching materials, exercise descriptions and performance records. In this best practice collection, we address this uncertainty and have summarized our recommendations and some of our best practice examples. Lecturers are welcome to adapt them to their teaching context! The examples and the course materials were developed for classes at the Bachelor’s and Master’s level at D-USYS of ETH Zürich. The main focus was on testing and implementing GenAI for information management and scientific writing in the classes in a meaningful way.
  • Analyzing argumentation patterns in political discourse for better policy design
    Item type: Journal Article
    Sudau, Manuel; Grêt-Regamey, Adrienne (2024)
    Journal of Environmental Policy & Planning
    A new policy is needed to manage Switzerland's increasing urbanization and growing population. The second stage of the revision of the Swiss Spatial Planning Law, which has been ongoing since 2014, aims to create the legal framework to reduce or prevent the negative consequences of soil sealing, such as loss of biodiversity or urban sprawl. The revision process is characterized by substantial opposition among the actors involved; an acceptable draft revision is not conceivable. Using a structuring qualitative discourse analysis, we coded the consultation responses from the three consultation processes to date. We based our code system on acceptance factors existing in the literature and analyzed the content of the consultation responses over time and by actor. The results show that while a consensus on the instruments of the policy is emerging, there is great disagreement about the exact design and the resulting effects on the actors involved. The relative advantage of the policy and its compatibility with existing regulations are not sufficiently elaborated and presented in a comprehensible way. Ultimately, we identify several patterns of argumentation that should be considered by the policy-makers involved in the further revision, especially to address the critical arguments of the cantons and municipalities. HighlightsPoints of conflict and consensus during policy design were identified using QDA.Structuring the content of the policy discourse can support consensus-building.Typical argumentation patterns for opposing or accepting a policy were identified.Growing consensus entails shifts of actor conflicts from mid- to low-level topics.Relative advantage of a policy solution is key to increasing its acceptability.
  • Application of Q-methodology for identifying factors of acceptance of spatial planning instruments
    Item type: Journal Article
    Sudau, Manuel; Celio, Enrico; Grêt-Regamey, Adrienne (2023)
    Journal of Environmental Planning and Management
    Worldwide, urbanization leads to increased pressure on prime agricultural land with irreversible impacts on the provision of life-supporting services such as food and drinking water production or habitat for plants and animals. As a basis for designing new policy instruments to protect soil resources, we applied Q-methodology to assess factors that influence the acceptance or rejection of such instruments. Using an online survey and interviews, we identified different social perspectives and their respective argumentation patterns. The results show that effect on people, institutional embeddedness, trust in the acting institutions, and the overall understanding of the instrument are the most important factors for the acceptance of policy instruments fostering the sustainable use of soil resources. During the interviews, idealistic and fact-based arguments were more important than person-based arguments. Based on our results, communication strategies in the policy-making process can be improved and tailored to the identified characteristics of the social perspectives.
  • Cases for Research Integrity: Generative AI
    Item type: Educational Material
    Paschke, Melanie; Mihálka, Réka; Sudau, Manuel (2024)
    Generative AI in research has triggered an ethical debate on how it uses might challenge long established codes and practices of research integrity. For example, a survey by Nature (Van Noorden and Perkel, 2023) of 1600 researchers highlighted among others the following topic of concern: Generative AI in research has the potential to introduce mistakes or inaccuracies into research texts, make the detection of fabrication or falsification of research more difficult or make plagiarism easier. It can bring biases into literature searches or embed bias or inequities into research texts. Other challenges when working with AI-based tools is the originality and the credits that can be given to the humans in the process, the accountability they have and the breach of IPR and copyright that is lurking in the shadows when using AI-based tools for creating content. Additionally, who owns the content done by AI? The creators that interact with the generative AI or the tech companies behind the algorithm? Confidentiality and privacy might be at stake when sensible research data is used in new AI-assisted workflows and training on the correct workflows for any personal, private, or sensible information becomes even more necessary. For sure, research processes as analyzing and publishing research data will change and generative AI will become embedded in the workflows. This raises questions around the energy consumption and carbon footprint of research. There is increasing need for students to discuss and practice research integrity with generative AI embedded and used in different research situations. In this collection, we provide case studies for moral reasoning and complementary material that can be easily adapted by lecturers to their research integrity classes.
  • Acceptance of Spatial Planning Policies for the Management of Soil Resources
    Item type: Doctoral Thesis
    Sudau, Manuel (2023)
  • Teaching Collection: Exercises and hands-on examples for ethical use of generative AI
    Item type: Educational Material
    Paschke, Melanie; Petterson, Alexander; Mihálka, Réka; et al. (2024)
    This collection of excercises is especially targeted to lecturers that want to teach their students to use generative AI in a responsible way with a special focus on responsible prompting.
Publications 1 - 7 of 7