Imagine a programming textbook that talks back. That’s the promise of EDBooks, a new platform designed to revolutionize programming education. Developed by researchers at the University of Michigan, EDBooks blend interactive narratives with the power of AI to create a more engaging and personalized learning experience. Traditional programming textbooks can be dry and static. It’s often hard to know which questions to ask, and even harder to get personalized feedback. EDBooks tackle this by letting learners interact with a virtual instructor through branching dialogues, posing questions, and receiving tailored responses. Think of it like a choose-your-own-adventure book for code. But what truly sets EDBooks apart is the integration of Large Language Models (LLMs) like ChatGPT. At any point, learners can step off the pre-defined path and ask their own questions, receiving contextually relevant answers from the AI. This blend of structured learning and open-ended exploration helps guide learners towards key concepts while encouraging their curiosity. EDBooks go beyond just text. The platform incorporates hands-on coding exercises directly within the narrative, allowing learners to test their understanding in real-time. Code examples are built up incrementally, step-by-step, with visual cues connecting the code to the explanations. This helps make even complex programs easier to grasp. A user study with 20 programming learners showed that EDBooks led to increased engagement compared to traditional learning materials and standard LLM UIs. Participants spent more time exploring the content and trying out the interactive exercises. While EDBooks show great potential, the research team acknowledges the need for improvements. Future work will focus on refining navigation, encouraging more meaningful exploration, and making the authoring process less manual. Could AI-powered interactive narratives like EDBooks be the key to unlocking a new era of programming education? The research suggests it’s a path worth exploring.