Write First, Then Draw a Card
Featured faculty: Anne van de Ven-Moloney
Program Director, Nanomedicine MS and Certificate Programs and Teaching Professor
College of Science
TL;DR: Draw an industry. Draw a technology. Build a startup with AI. Then draw an audience nobody planned for.
What she’s doing: Anne teaches Nano Biomedical Commercialization: From Bench to Bedside, a graduate course in which student teams develop a small business proposal over the semester. Each week brings a new writing prompt, a team response, feedback, and revision. From the first class, Anne explained why AI would stay out of those assignments: the course is built to develop professional skills for industry, and students were receptive.
She did not want to ignore AI entirely, so she held it until the end. After her students had written and received feedback on each section of their proposals, she ran one gamified session that took students back through the same stages with AI as a thinking partner. After discussing how they already use AI, the class brainstormed where it might help a proposal, from strategic positioning and data gathering to stakeholder perspectives and polish. Pairs then drew an industry card (e.g., veterinary medicine) and an emerging technology card (e.g., 3D bioprinting). In timed rounds driven by one-sentence prompts on the slides, students worked in pairs to ask AI for five commercialization ideas, chose one that connected to their own experience, developed a business plan, had AI critique it as investors or regulators might, generated pitch keywords, and wrote a one-minute investor pitch. A final card named an unconventional audience, such as sixth graders touring the company or lawyers weighing a lawsuit, and teams rebuilt the pitch and presented both versions.
What’s working and what isn’t: Anne encouraged pairs to use two AI tools side by side, and everyone did. Claude and ChatGPT often answered differently, and for some steps one was clearly stronger, so comparison became part of the lesson. The stakeholder critique targeted a problem she sees often: students steeped in the science write for scientists, not investors. By the end, students could ask AI to read their own proposal as an investor would while keeping their voice. Because students do the experimenting and report back, the instructor need not be an AI expert; a colleague new to AI in the classroom helped run this session comfortably.
Timing was the main friction. Debriefing after every step was too much for one session, and Anne plans to group two or three AI steps between discussions. She considered moving the activity earlier but kept it late, since the pull toward AI as a shortcut remains strong. She also notes that hidden AI use now seems to be the most common source of conflict she sees in team projects, so her syllabus asks teams to be transparent from day one.
| Adapting Across Contexts: Two design choices carry this activity. Students write first, so AI arrives as a reviser rather than an author. The cards impose constraints that nobody chooses. This randomness prevents students from settling on a comfortable idea and makes the AI’s output something to judge rather than accept. The audience swap at the end tests whether students understand their idea well enough to explain it to anyone. The cards are the most portable piece. A law course might deal a client and a jurisdiction, then a skeptical jury. A journalism course could draw a beat and a data source, then require the story be retold for a community radio call-in show. An architecture studio might pair a site with a material, then present to neighbors opposed to construction. Large sections may need fewer whole-class debriefs, and online courses could run the rounds asynchronously. |
What’s next: Nearly every one of Anne’s assignments now includes AI tips, framed around students doing the work first and keeping their own voice. The tips mostly work, though some could be phrased better. When student work shows her a prompt that needs tightening or another question, she edits it in Canvas immediately rather than trusting a note to survive until next year.
| Try it yourself! Anne has shared her class slides and printable cards. The deck includes each round’s prompt, the closing discussion questions, a summary slide of practical ways to use AI on a business proposal, a short list of do’s and don’ts, and guidance on citing AI. |
Have you tried something new with AI? How did it go? Send an email to [email protected] to let us know!