Appendix D — Generating Course Material with Gemini Notebook
This appendix describes the use of Gemini Notebook (NotebookLM) as a tool supporting the preparation of course material, complementing the use of AI for reviewing text and code, mentioned in this book’s preface.
D.1 Conceptual videos and EP-solving videos
For each chapter, a project was created in Gemini Notebook using the full PDF of the book and the morph.py file as sources. From these sources, Gemini Notebook automatically generated a short video (7 minutes on average), used at the start of classes. Chapters alternated between two types of video:
- Conceptual video, presenting the chapter’s theoretical foundations, generally shown in the first class dedicated to the topic;
- EP-solving video, presenting a solution strategy for the Programming Exercises, shown in the following class, when most EPs were solved together with the class.
The video generated by Gemini Notebook does not replace the instructor’s explanation. It works as a brief introduction and motivation for the topic, giving the student context before the slides are presented and the Colab notebooks are opened.
D.2 Supporting slides
Complementing the videos, Gemini Notebook also generated, from the same sources (the book’s PDF and the morph.py file), a set of approximately 12 slides per chapter, covering both the theoretical concepts and the practical part. Generation used a chapter-specific prompt, restricting the scope of content to be synthesized and avoiding both anticipating topics from later chapters and reproducing long passages of the book without didactic adaptation.
After the slides were presented, class continued with the opening of the chapter’s two Colab notebooks (theoretical and practical), emphasizing the interactive simulators and running the code blocks with parameter changes.
D.3 Main material and final remarks
Generating videos and slides with Gemini Notebook, based on the book itself and the morph.py library, made it possible to standardize the opening of classes and reduce the time spent preparing course material, without dispensing with the instructor’s curation in crafting the prompts and conducting classroom activities.
The videos and slides produced by Gemini Notebook are motivational in nature and serve as a starting point for discussing the concepts presented in each chapter. Like any AI-generated content, they may contain occasional inaccuracies or errors. In some cases, these inaccuracies are intentionally exploited during class to check whether students are critically following the presentation and can identify conceptual inconsistencies.
The course’s official content, however, is the material produced by the method described in this book’s preface, made available in three complementary formats: HTML, for browsing with interactive simulators; PDF, for reading and printing; and Jupyter notebooks (.ipynb), for running and experimenting with the code. All material generated by Gemini Notebook should be understood as a complement to this main content.
As with the other uses of generative AI described in this book, responsibility for the technical quality, conceptual correctness, and pedagogical adequacy of the material remains with the instructor.