What we'll be doing over the next few years will involve specific adaptations to the strengths and weaknesses of specific tools, but the underlying change in teaching and learning is moving faculty towards being more explicit about both goals and strategies to pursue those goals in their classrooms.
Generative AI in Education
The educational technology landscape is evolving quickly. Generative AI has enabled a large number of new technological capabilities that alter student learning and educational approaches. Across NYU’s global network there are a variety of resources for NYU Abu Dhabi faculty, instructors, and students as we adapt to AI-enabled tools.
Teaching and Learning resources from NYU’s Global Network are available on the following websites:
Video: Generative AI in the Classroom
Generative artificial intelligence, which includes natural language processing systems, may have disrupted some of the traditional classroom practices but it also offers new and exciting opportunities. The prevalence of generative AI demands a rethinking of traditional assessment methods as well as a call to come up with creative ways to use it for an enhanced learning experience. Watch this video to learn more about how you can adapt your teaching style to take Generative AI into account.
Openai’s ChatGPT4 (version launched March 14th, 2023) is a generative AI platform that produces text, code, and images (through Dall-e) from simple prompts.
This new generative AI platform, and others like it are being used by students. It is a powerful new tool for generating sophisticated content. We need to respond to it in our courses, especially as the sophistication of the platform’s capabilities is constantly increasing.
Below are some suggestions of how to talk with your students about the rules and regulations of using this tool in your fall 2023 courses. You can get a free account at chat.openai.com.
How to talk with your students about the rules and regulations of using AI:
It is crucial to note that you should never enter personal student information into ChatGPT, nor your own. The platform is trained on what you enter into it and the content you enter in can be reproduced in a response to someone else’s prompt.**
We ask all faculty to include a ChatGPT statement on their syllabus and discuss this with your students. Please also ensure you have a notice on your syllabus that you reserve the right to use TurnItIn for written assignments.
To get started in reflection on how ChatGPT impacts your curriculum and pedagogy ask yourself these four questions:
- What are the cognitive tasks students need to perform without AI assistance? When should students rely on AI assistance from ChatGPT or similar platforms (Google’s Bard, Microsoft Bing, etc.)?
- Where can an AI facilitate a better outcome?
- Are there efficiencies in grading that can be gained? Are new rubrics and assignment descriptions needed?
- Will you add a generative ai code of conduct to your syllabus?
Whatever strategy you choose to pursue, all strategies require explicit communications with your students about your expectations and your hopes for what they will learn from the assignments. Please include a description of your preference of when, how, and where ChatGPT can be used in your course on your syllabus and discuss it with your students.
NotebookLM
NotebookLM is an AI-powered research assistant designed to help students and faculty synthesize information and engage with reading materials in new ways. Unlike ChatGPT, which generates output based on knowledge from pre-trained datasets, NotebookLM allows users to easily organize and manage course materials by uploading files (PDFs, websites, text, videos, or audio files) and generating responses based on the uploaded sources.
A key feature of NotebookLM is its citation system — whenever it provides information, it references the specific document and passage it was drawn from, ensuring transparency and reliability. Though it is always good to verify when using generative AI tools.
With the Deep Dive Audio feature, NotebookLM can also create podcasts based on reading materials that summarize dense material. This feature makes academic content more accessible and adaptable to different learning preferences and can be used, with faculty supervision, to enhance comprehension and engagement with the materials. It can also produce these podcasts in multiple languages, increasing access to new disciplines for non-native English speakers.
How to get started with NotebookLM?
1. Sign in to NotebookLM using your Google account. On the main page, click on “Try NotebookLM,” which will direct you to the page where you can create a new notebook.
2. Create a new notebook, where you upload different documents (lecture materials, research papers, textbooks, etc.) and interact with them through a chatbot.
3. After uploading the sources, you can interact with their documents in various ways, such as analyzing key information, generating discussion questions, summarizing content, and generating a “Deep Dive” podcast conversation.
Using NotebookLM as a Pedagogical Tool
-
Upload textbooks, research articles, or lecture notes, and use NotebookLM to generate structured summaries that highlight key concepts and main ideas.
Figure 5. An example prompt to summarize lecture notes -
NotebookLM can generate reading guides with annotations and key takeaways to help students navigate complex material. It can also create reading comprehension and discussion questions to encourage deeper engagement with the text.
Faculty can use it to highlight key themes, clarify difficult concepts, and connect readings to broader course topics.
However, while NotebookLM has a relatively low hallucination rate, it’s important to manually review the summaries and questions it generates to ensure accuracy and relevance.
Figure 6. An example prompt to generate a reading guide of the book. -
Based on uploaded documents, NotebookLM can suggest questions tailored to specific learning objectives and create quizzes, case studies, or practical exercises.
Figure 7. A sample prompt to generate a quiz based on the lecture materials. -
NotebookLM makes it easier to manage course materials and streamline teaching by:
- Finding Key Passages Quickly: With uploaded textbooks, research papers, and lecture notes, faculty can quickly search for specific quotes, key arguments, or definitions when preparing for class or responding to student questions.
- Keeping Course Materials Organized: In NotebookLM, faculty can organize readings by topic, week, or theme, making it easier to access relevant materials when planning lectures or discussions.
- Providing Context for Student Questions: When students ask about a concept, faculty can retrieve relevant excerpts from course materials to support clear, accurate explanations.
- Supporting Office Hours and Student Feedback: Faculty can store student questions and past discussions in a notebook, allowing for more personalized follow-ups and helping students connect ideas over time.
Figure 8. An example prompt to extract specific quotes or mentions from the uploaded materials. Figure 9. An example prompt to organize materials based on key topics. Figure 10. An example prompt to generate citations.
Guidelines for Using NotebookLM in Research and Teaching
While NotebookLM can be a valuable tool for organizing and engaging with course materials, it is important to approach its use with careful consideration.
- Organization and Citation Management: It is important to verify that the citations generated align with the required citation style, as automated tools may occasionally require adjustments.
- Ensuring Accuracy and Reliability: While NotebookLM has a relatively low hallucination rate, AI-generated content should always be reviewed for accuracy. Faculty are encouraged to cross-check summaries and citations against original sources before integrating them into their work.
- Guiding Students in Responsible AI Use: As AI becomes more accessible, students may rely on it in ways that do not align with the learning goals. Faculty can help students understand how to engage with AI tools thoughtfully, emphasizing their role in supporting comprehension rather than replacing independent analysis.
By using NotebookLM intentionally and within academic guidelines, you can explore its potential benefits while maintaining high standards of scholarly integrity.
Enhancing Accessibility and Engagement with Deep Dive Audio Podcast Feature
Traditional academic readings may at times be challenging for neurodiverse learners and students from diverse linguistic and academic backgrounds (for instance, students who have studied in languages other than English prior to coming to NYUAD). Using the Deep Dive Audio feature can enhance the accessibility of course readings and increase student engagement. Faculty can use it as a supplement for traditional reading assignments by generating audio versions of key readings and turning them into personalized podcasts.
Generating AI Podcasts with NotebookLM
-
- Upload readings or course materials to NotebookLM.
- The AI converts the text into a structured, spoken format, ensuring a natural, engaging listening experience.
- You can prompt the software to customize the podcast by focusing on specific themes or key concepts to align the audio content with the course objectives.
- Review the AI-generated transcript to verify accuracy and ensure no AI hallucinations before sharing it with students, and ensuring that all students have access to the same reliable version of the material while still allowing space for thoughtful interpretation and discussion.
- Specify to students how they should engage with the AI-generated material and what might be the advantages or drawbacks of incorporating it in their learning. If you wish students to still engage with the original text, despite having access to the podcast, explain why you are asking them to do that.
Sample Use
"Create a podcast on the three readings for Mind, Matter, and the Machine, aligning with the syllabus and course learning outcomes. Ensure students grasp key ideas and can answer:
(1) What is the basic result of the neuroscientific experiment?
(2) How might one resist the idea that the experiment challenges free will?-"
NotebookLM - Deep Dive Audio Podcast review. In a few minutes, NotebookLM generates a short podcast, which you can listen to in order to fact-check for any hallucinations. Once you are sure that it fits the requirements, you can upload it as an audio file on Brightspace for students to listen to.
NotebookLM - Podast, upload it as an audio file to Brightspace. You can also share a link to the project directly to students:
NotebookLM - share a link to the project directly with students. -
NotebookLM - Interactive Learning with Real-Time Conversation NotebookLM Interactive NotebookLM’s Interactive Mode offers a way for students to engage with course materials by allowing them to join live conversations between the podcast hosts. This feature enables students to become active participants, asking questions and discussing key topics as they arise during the podcast.
It’s particularly useful after reading assigned texts, helping students clarify confusing points or deepen their understanding through real-time discussions. Even while listening, learners can pause and contribute questions that reflect their thought processes, making the experience dynamic and responsive. This approach introduces a new, interactive form of learning — like having a personal tutor who guides students through complex material, simulating a conversation with the professor. The feature transforms passive listening into an engaging learning session, offering a unique opportunity for students to explore course content collaboratively.
-
NotebookLM’s Deep Dive Audio feature promotes inclusive teaching by supporting diverse learning preferences and processing styles. Research shows that students engage with instructional materials in multiple ways — some retain information more effectively through reading, while others benefit from audio formats that allow for multitasking or accommodate different cognitive needs (CAST, 2018; Fleming & Mills, 1992).
By offering both reading and audio options, NotebookLM enables learners to engage with content in the format that best supports their individual needs. This dual-modality approach aligns with Universal Design for Learning (UDL) principles, which emphasize providing multiple means of representation to enhance accessibility and comprehension (CAST, 2018).
Supporting Neurodivergent Learners
The Deep Dive Audio feature is particularly beneficial for neurodivergent students. Students with ADHD, dyslexia, or other cognitive processing differences may find large blocks of text overwhelming or difficult to process. Research suggests that audio-based learning can reduce cognitive load and increase comprehension for these learners (Guerin et al., 2021; Schwanke et al., 2020).
By offering course content in audio form, NotebookLM helps reduce the strain associated with traditional reading, making learning more approachable and less taxing for students who benefit from auditory processing or need to move at their own pace.
Supporting Flexible Learning
Audio delivery also meets the needs of students navigating complex schedules. Students can listen to Deep Dive summaries while commuting, exercising, or completing routine tasks — helping them stay connected to course content even when sitting down to read may not be feasible. Flexible access to materials in varied formats has been shown to increase student engagement and learning outcomes (Mayer, 2009).
-
A common concern among faculty is that tools like NotebookLM’s Deep Dive Audio feature may stop students from developing critical reading skills. Reading carefully and critically is undeniably a valuable skill that educators want their students to acquire. Therefore, it’s crucial to remember that NotebookLM should be used to complement, not replace, reading.
A Supplement, Not a Substitute: Remind students and teach them that the audio feature serves as a supplementary tool. Students can listen to the podcast to get an initial overview, helping them grasp the main ideas before diving into the text. But reading the text themselves is crucial to student learning. Alternatively, they can read first and then listen to reinforce their understanding, or even skim the reading and use the audio to fill in comprehension gaps.
Supporting Deeper Engagement: Audio learning can act as a primer that encourages more engaged reading. By familiarizing themselves with key concepts through listening, students may approach readings with a clearer sense of purpose, focusing on analysis rather than initial comprehension.
Interactive Discussions Reinforce Critical Thinking: After either reading or listening, students can participate in the Interactive Mode discussions. This dynamic engagement ensures that the material is not passively consumed but actively processed, discussed, and questioned, thereby fostering critical thinking skills.
Changing Nature of Reading in the Age of AI: While traditional reading skills remain important, the ways in which we engage with content are evolving. The ability to synthesize information from various formats — including audio — is becoming essential. Far from diminishing reading skills, tools like NotebookLM equip students with adaptable strategies for engaging with complex material.
Sources
CAST. (2018). Universal Design for Learning Guidelines version 2.2. Retrieved from https://udlguidelines.cast.org
Fleming, N. D., & Mills, C. (1992). Not Another Inventory, Rather a Catalyst for Reflection. To Improve the Academy, 11(1), 137–155.
Guerin, A., Murphy, P., & Murray, S. (2021). Supporting Neurodiverse Learners in Higher Education. Inclusive Practice Reports, 3(2).
Schwanke, T. D., Smith, R. O., & Edyburn, D. L. (2020). "Assistive Technology and Universal Design for Learning: Two Sides of the Same Coin." Journal of Special Education Technology, 35(2), 95–104.
Mayer, R. E. (2009). Multimedia Learning. Cambridge University Press.