A few days ago, David Wiley, Chief Academic Officer of Lumen Learning, posted a blog article titled Generative Textbooks. After a short introduction, he wrote that he “started wondering – what if, in the future, educators didn’t write textbooks at all? What if, instead, we only wrote structured collections of highly crafted prompts? In linear fashion, the learner would use the prompts to interact with a large language model (LLM).”
Dr. Wiley elaborated on his vision and predicted that “many students will quickly abandon the traditional format textbook in favor of the more interactive, open-ended, and personalized nature of the generative textbook.”
I forwarded David’s article to several friends. One of my friends is a major tech and learning geek. During our discussion, I asked him what he thought of the idea that you could take the generative textbook idea a step further and utilize an app integrated with the LLM to create a course and transfer the tested competencies of the student to a digital learning record. All possibly completed without the aid of an LMS. While we agreed the concept was probably possible, we didn’t derive a neat solution (yet).
Evidently, enough people commented and replied to Dr. Wiley’s post on LinkedIn, that he wrote another article titled Generative Textbooks – A Brief Example that outlined his concept of the structured collection of highly crafted prompts. It’s worth reading, particularly if you’re a teacher weighing your options of how to utilize LLMs in your classes this fall.
I’m not sure if the examples of structured prompts created by David Wiley are too detailed/too numerous or not detailed enough/too few. It’s a great list based on the subject matter. I think the answers to those questions are something that a teacher/instructor would have to experiment with. I also don’t know if his intention is for each student to utilize every prompt or just the ones that reflect the subject area that they want to review in more detail. The answer may depend on whether a traditional textbook is used along with the “generative textbook.”
His instructions to scan a page from an open-source textbook that he used to derive his prompt content led me to believe that his intention was to provide the specific content he used to generate the prompts. At the same time, I believe it’s possible to scan text into an LLM and use it as a more directed source for generating answers to prompts that are less likely to experience incorrect information or response hallucinations (Nolej is a company that has a product that does something just like that for generating learning modules).
I returned to the first article/post of Dr. Wiley to look for the following paragraph:
Imagine how much more natural it would be to teach metacognitive skills, information literacy, and related topics when a learner’s primary activity is asking questions of an LLM, rather than reading a static text. Learning to ask useful questions – whether of an LLM, another person, or the universe itself – is directly at the center of the educational enterprise.
I love the point that Dr. Wiley makes in this paragraph. We all know that most students at some point will utilize LLMs to review or assist in the preparation of homework. At what point in their learning journey do you provide them with fewer and less specific prompts and allow them to experiment (and grade them accordingly for their creativity or ingenuity)?
Another question that surfaced as I reread and thought through both of Dr. Wiley’s posts was If every student in a class of 15 enters the exact prompts specified by the instructor, will the answers be the same? I don’t know the answer to that question. I also don’t know if some students would regenerate their responses to see if the LLM gave them a slightly different response. I guess it might depend on whether the teacher/department/university was using a LLM with “guardrails” that tightly controlled the source material.
The final paragraph of Dr. Wiley’s first post ended with:
In the case of generative textbooks, prompt engineering is learning engineering, or instructional design, or whatever they call it in your world.
It is clear that his intended audience was those of us who teach. If you’re ready to increase your learning, the Generative AI era may be the most exciting and challenging era for making sure your students learn the course material and learn to utilize the AI tools that will guide them in their academic and professional journeys.