Nicole LaDue
Professor
Department of Earth, Atmosphere and Environment
College of Liberal Arts and Sciences
Course(s) Targeted by this Innovation
EAE 120: Planet Earth
Development and Implementation
2026-2027
Purpose and Impact
EAE 120: Planet Earth is currently only offered in an online, 8-week long format. It attracts a large number of students in the Bachelors of General Studies, those filling general education credits for their degrees, and non-traditional students who are not able to attend class in person. The course currently reaches 250 students per semester, and the EAE department is considering increasing enrollment.
Since this is an online course, we have a large number of students who engage in mis-use of AI (e.g., copy and pasting assessment questions into AI, using AI to generate discussion board posts). The course was designed with best practices for online pedagogy, but many of those techniques are not working well because students are using AI.
Lastly, this is an important candidate for redesign because it is an online science class. There are a range of tools we are using that may not be heavily used in other classes (e.g., drawing/sketching diagrams, labeling diagrams, arguing with evidence).
Description of Innovation
First, I will redesign the discussion board activities to build AI literacy. Specifically, students will have to investigate and discuss AI hallucinations. It is important to help students understand how AI generates answers and the types of errors that AI make. Next, students will need to prompt various AI models to determine whether it can effectively label a science concept as a theory or a hypothesis. This builds science literacy as well as AI literacy. I would also like to include an activity for students to learn more about the environmental impact of AI and Bitcoin mining, which addresses the ethics side of responsible AI use in the context of the content of the course.
Additionally, in previous semesters I have asked students to generate sketches. The types of sketches AI generates have very entertaining hallucinations. Students can engage in critical thinking about the content to identify and correct the errors. This builds AI literacy as well as critical thinking skills. Currently, I use a labeling activity where students download a photo and label the fault in the photo. AI models are able do this task, but cannot do it correctly. This type of diagram-based activity may require shifts over time as AI tools become better trained to identify diagrams. Currently it is the only type of activity for which AI is ineffective at generating plausible answers.