Concepts for Teaching Software Development in the Age of AI-Tools

LLM-based tools such as ChatGPT, GitHub Copilot, and the like have already arrived in software development practice, and continue to change the way of how software is created. Students use Generative AI tools for any purpose, whether we agree with the use or not. In this paper, we argue that in order to appropriately prepare our students for professional life, we educators need to incorporate the use of these tools explicitly into the teaching and learning process, so that students will learn a systematic and professional usage of these tools. Teaching basic software development concepts to students with only little or no prior knowledge has always been a challenge. Incorporating GenAI-based tools into software development education requires some careful rethinking of teaching concepts, especially with respect to constructive alignment, i. e. aligning learning objectives, teaching and learning methods, and exams to address GenAI-based support. Based upon first observations on the usage of GenAI tools in class, in this paper we suggest and compare concepts for teaching and learning basic software development that integrate GenAI tools at different levels of intensity and at different points during a semester. From these observations, we derive recommendations both for teaching and learning settings and for corresponding assessments, for teaching software development in a constructively aligned way in the age of GenAI tools.