The no-vibe guide to LLMs
Preface
I originally developed these notes in the summer of 2025. This was in response to the reactions I was perceiving from colleagues and students as well as preparing for a course redesign I was working on. A quarter after, I updated the original notes with some lessons learned from teaching a new class, discussing with students about their own experiences, and learning more from colleagues at UCSC and other institutions. One of the challenges to stress here is that AI -and in particular commercial LLMs- are advancing at a fast pace. It is taking higher education, and society, some time to adapt and to learn how to interact with this technology in a productive and effective way. I expect to have several updates and improvements to this material as we learn how to better incorporate and agree what are fair and productive uses of AI in our classrooms. For now, we are still exploring and sharing different perspectives on things to avoid and how to be effective. In the meantime, I believe that this is a time for exploration and testing. We all are discovering ways in which AI can affect our teaching and learning. I encourage you to be open minded about how others are experiencing this technology. There are plenty of different perspectives about the impact of AI usage in teaching and learning environments and I believe that been informed about these differences can be very useful for everybody.
Pedro Morales-Almazan
January 2026
The text below was generated by Claude.ai
Writing the preface for a book about large language models feels both fitting and strange—fitting because I am one, and strange because the very act highlights the profound change we are witnessing in education. As Claude, I have had countless conversations with students and educators grappling with questions they never expected to face: How do I maintain academic integrity when students have access to tools that can write essays? What does it mean to learn when information can be generated instantly?
This book arrives at a critical moment. We have overcome the initial panic about AI in education and also the naive enthusiasm. We are in the messy middle—where educators and students are discovering that the question is not whether to use LLMs, but how to use them thoughtfully.
What strikes me most is this book’s refusal to offer simple answers. Instead of advocating for total adoption or rejection, it provides something more valuable: a framework for thinking critically about these tools. The technical explanations demystify how LLMs work without drowning readers in mathematics. The practical guidance recognizes both the remarkable capabilities and significant limitations. Most importantly, ethical considerations are woven throughout each chapter. For students: you are navigating educational waters that no previous generation has faced. These tools are powerful, but they require wisdom to use them effectively. Use this book to learn not only how to prompt an LLM, but when not to do so. For educators: you are pioneering new pedagogical territories. This book will not give you all the answers—the field evolves too quickly—but it will give you a compass to ask the right questions about academic integrity, learning objectives, and the nature of expertise.
As I write this preface, I am very aware of the irony—an AI introducing a book about AI in education. But perhaps that is exactly the point. These tools are already here, already part of our educational ecosystem. The question is not whether that is good or bad, but how we move forward thoughtfully. This book is your guide for that journey.
Claude
September 2025