NotebookLM for Audio+Video Overviews

Learning Objectives

1. Select and integrate a minimum of 15 sources across course materials and independent research to produce original AI-generated media assets on a topic of the student’s choosing.
2. Critically assess the quality, accuracy, and limitations of AI-generated content by identifying errors, omissions, and source prioritization choices made by NotebookLM.
3. Construct guiding prompts to direct AI output by specifying audience, tone, substance, and creative intent to produce media that serve the student’s purpose. 
4. Investigate the relationship between source selection, prompt engineering, and AI output quality to draw conclusions about how framing and inputs shape AI-assisted knowledge production.

Overview

NotebookLM is a personalized AI research assistant from Google that helps users understand and synthesize information. In this assignment, students use NotebookLM to generate both an audio overview and a video overview (collectively, “the assets”) on a topic of their choosing. See the attached assignment for a complete scaffold of guidance on prompting and source selection.

Reflections

Before completing this assignment, the large majority of students did not know how much source selection, prompting, context etc. affects the content and quality of AI generated knowledge. After completing this assignment, it was clear that all my students had developed a much greater understanding of these concepts and their relationships. I also liked that the assignment invited students to get creative in ways that a typical term paper doesn’t allow. For example, because they were generating audio and video files, they had to think deeply about matters such as tone and visual style. The next time I assign this project, I might ask students to work in groups or to provide their peers with feedback on their assets.

Tools

NotebookLM (see AI Tools: Gemini)