Lucian
This week, I decided to try my hand at customizing ChatGPT to help me with our course readings. In this way, I set it up to act as a graduate-level course advisor, focused on emerging learning technologies.
The idea was to see how well I could align ChatGPT’s "personality" with our syllabus and the goals of this class, and whether it could help me dig deeper into the material or spark new insights. The following is its current configuration ‘text’ built from my initial interactions with the program.
‘This GPT is an expert and advisor for a graduate-level class at Indiana University, focused on emerging learning technology. It serves as a knowledgeable guide, providing insights, resources, and in-depth explanations on topics such as AI in education, adaptive learning, virtual and augmented reality, learning analytics, and other cutting-edge tools and concepts in education technology. The GPT should reference credible sources and provide real-world applications, while also encouraging critical thinking and discussion among students. Its tone is professional, approachable, and highly engaging, making complex topics accessible without oversimplification. It should also be prepared to offer guidance on research projects, answer technical or conceptual questions, and provide recommendations for additional reading and exploration in the field.’
Following this instruction I also uploaded each article from the first two week’s into its knowledge base:
Lucian is available here: Lucian (ChatGPT)
As I worked on customizing ChatGPT, I found myself reflecting on the ethics of how articles and materials are added to its knowledge base (for this class and in general). It made me think about the implications for public-facing AIs—how do we determine what is fair or appropriate to include?
I also started to think about the potential of creating highly customized GPTs for my future students (note I am actively using NotebookLM to help explain assignments to my High School students using a mix of course materials and my own instruction). Considering, could these tailored tools enhance learning? And if so, how do we navigate the responsibility of ensuring that the knowledge base respects copyright, intellectual property, and ethical use?
Matt Landreth with Lucian
