The introduction of generative artificial intelligence into study, synthesis and presentation workflows raises important questions about its educational potential. This article examines Google’s new NotebookLM presentation-generation feature, evaluating its performance through an experiment based on the Italian Ministry of Education and Merit’s guidelines on the use of AI in schools. The analysis shows that, although the tool cannot replace critical reading, it significantly enhances users’ ability to reorganize and communicate complex content.
1. Introduction
The growing integration of generative AI into productivity tools and document-analysis systems is changing how complex materials are synthesized and communicated. NotebookLM, developed by Google, recently introduced advanced functions for automatically transforming documents into infographics and structured presentations.
To assess its effectiveness in a real context, a highly detailed and relevant document was selected: the Ministry of Education and Merit’s guidelines on the introduction of AI in Italian schools. The document contains a rich set of pedagogical principles, governance requirements, professional roles and risk-mitigation strategies. Its density and complexity make it ideal for testing NotebookLM’s ability to extract, reorganize and present structured information.
2. Complexity of the MIM Guidelines
The MIM Guidelines include several key components:
-
an anthropocentric approach to AI, centered on human value and agency
-
a series of fundamental pillars, including safety, transparency, responsible use and skill development
-
a clear distinction among professional roles such as school leaders, administrative staff, teachers and students
-
a list of risks associated with AI adoption and corresponding mitigation measures
-
various application scenarios, from personalized learning to administrative automation and inclusive education
The diversity of conceptual layers makes the document challenging to summarize and transform into a clear visual format.
3. Using NotebookLM for content transformation
The experiment consisted of uploading the official guidelines PDF into NotebookLM. The system immediately indexed the key sections and activated its synthesis features, question generation and—most notably—the automatic presentation builder.
The resulting presentation displayed several noteworthy qualities:
-
effective structural organization: the tool identified the main pillars and grouped them into coherent slides
-
reduction of complexity: dense concepts were broken down into accessible, digestible units
-
neutral and consistent style: no interpretative distortions, just clear reformulation
-
visual responsiveness: lists and tables were converted into clean graphical elements suitable for sharing and teaching
Visually and conceptually, the output respected the integrity of the source while improving readability.
4. Methodological analysis
4.1 Document decomposition and indexing
NotebookLM using semantic parsing breaks the text into conceptual clusters, detecting headings, subsections, lists, keywords and role descriptions.
In the case of the MIM Guidelines, it correctly identified:
-
main chapters
-
descriptions of professional roles
-
lists of risks and opportunities
-
ethical and pedagogical principles
-
final recommendations
This structural understanding is the foundation for the generated presentation.
4.2 Presentation generation
The generated slide deck followed a coherent narrative pattern:
-
introductory context slide
-
structural overview of pillars and stakeholders
-
thematic slides on risks, principles and practices
-
a concluding slide on responsible AI adoption
The organization shows that the system is able not only to summarize but also to construct a usable communicative pathway.
5. Critical discussion
Despite its strengths, the tool has limitations:
-
no conceptual interpretation: NotebookLM does not analyze implications or policy depth
-
risk of oversimplification: intricate concepts may become too linear
-
total dependence on the source: omissions or ambiguities in the input remain in the output
-
need for expert oversight: neutrality alone does not guarantee correctness
Still, the speed and clarity with which it reorganizes the material is remarkable.
6. Educational and professional implications
This experiment suggests that tools like NotebookLM can have a significant impact across educational and professional contexts:
-
teaching: fast creation of lessons, course materials and visual supports
-
research: accelerated reading, comparison and summarization of dense documents
-
school governance: clearer understanding of regulations and guidelines
-
institutional communication: rapid production of accessible materials
The tool acts as an amplifier, not a replacement, of human expertise.
7. Conclusion
NotebookLM’s new automated presentation feature represents an important step in the integration of generative AI into knowledge-management workflows.
The experiment conducted using the MIM Guidelines shows that the tool can reorganize complex materials into clear, communicative formats while preserving conceptual coherence.
AI does not replace critical reading here—it enhances it.
NotebookLM emerges as a valuable resource for educators, researchers and professionals, opening new possibilities for working with complex documents.
I’m sharing the document generated with NotebookLM