Thursday, April 16, 2026 — An open dialogue about how AI is showing up in teaching across campus. Follow-up to Asa Stone's PRAIRIE Seminar on the Taxonomy of AI Integration.
| Time | Session | Location |
|---|---|---|
| 1:00 – 3:00 PM | From Shared Language to Shared Practice | Kiewit Hall A310 |
An open conversation for faculty, students, and staff who are curious, cautious, or already experimenting with AI in their courses. No prior AI experience required. Come with questions, concerns, or ideas.
A shared language is within reach. Derek Heeren presented a six-level AI use taxonomy (No AI → Tool → Tutor → Collaborator → Co-Creator → Agent) drawn from Asa Stone's March 31 seminar and informed by all three Prairie advisory boards. Faculty naturally mapped their demonstrations onto these levels, suggesting the framework is intuitive and ready for refinement through continued CoP conversation.
Process over product is already happening. Both CS presenters independently arrived at process-over-product assessment models. Stephanie Valentine built tiered AI permissions into Cursor using rule files and SpecStory conversation logging. Alisha Bevins eliminated artifact grading entirely, replacing it with five discussion-based checkpoints where students demonstrate understanding verbally to TAs. "AI is wonderful at making an 80% solution. The last 20% is based on expertise and skills we've been teaching."
Most classrooms are stuck at the tutor level, but industry expects more. Heidi (engineering education) raised the central question: are we lagging behind? Mona Bavarian confirmed that companies like Dow, Exxon, and Chevron want collaborator-level fluency, and some now include AI usage in performance evaluations. The gap between classroom practice and industry expectation is real and urgent.
Discipline-specific integration is emerging organically. Chemical engineering (polymer prediction tools, materials discovery agents), computer science (structured AI coding environments), education (rural AI curricula, art education chatbots), and environmental engineering (AI for sustainability communication) each brought distinct applications. The shared language needs to be flexible enough to hold all of them.
Nebraska's K-12 AI gap. Minji Jeon highlighted that Nebraska currently has no AI standards, graduation requirements, or professional development funding, while states like West Virginia, Tennessee, and Louisiana are moving ahead.
Nirupam Aich showed how ChatGPT graphics evolved over one year, from an overcomplicated research diagram to a freshman-friendly visual: problem → tools → solution. His slogan: "Science + Engineering + Data = Solutions for People and the Planet." Then he flipped the script: what about the chemical pollution footprint of AI itself? Data centers aren't clouds. They're material-intensive physical systems with HVAC emissions, cooling fluid carcinogens, and semiconductor metals leaching into water.
Alisha Bevins opened with a provocation from a GitHub Education engineer: "Everyone said everyone should learn to code. Now everyone can code. So what does the future look like?" She noted he hadn't written syntax in months but was still exercising deep CS skills. Her TAs were "bored out of their mind" grading artifacts, and are now energized by verbal checkpoint discussions.
Stephanie Valentine's Cursor demo showed how rule files can turn an AI coding assistant into a pedagogically governed tutor. She worked directly with the SpecStory developers to add conversation logging, then required students to submit their full AI interaction history alongside their code.
Mona Bavarian offered live access to CopredPred, her lab's copolymer prediction tool (password: "Husker AI Days 2026"), and invited attendees to test it against published reactivity ratios.
Minji Jeon connected Nebraska's K-12 AI policy gap to the higher ed opportunity. While West Virginia, Tennessee, and Louisiana advance state AI standards, graduation requirements, and professional development funding, Nebraska has none in place. She pointed to the taxonomy paper she co-authored with Justin Olmanson and Azadeh Hassani, which defines six categories of student-AI engagement to help faculty anticipate how students will use AI across assignments and plan for learning-centered interactions. The six-category structure paralleled Derek's six-level classroom taxonomy from earlier in the session.