CCSC:MW 2026 · Three-Hour Workshop
Designing AI-Constrained Collaborative Activities
for Computing Education Using coLearn-AI
A hands-on workshop for computing and other educators who want to use AI in ways that preserve student reasoning instead of replacing it.
James Paul Skon · Kenyon College / Karolinska Institutet · skonjp@kenyon.edu
Open this site on your laptop: tinyurl.com/colearnai
Explore coLearn-AI
Try the demo instance on your own, read the project overview, or go to the Resources page for papers and workshop materials.
coLearn-AI Research
Research on collaborative learning, computing education, and AI-mediated learning environments: projects, publications, and the people behind them.
What This Workshop Is About
Large Language Models make it possible for students to generate correct answers and working code while engaging minimally with the reasoning those outputs are intended to demonstrate. We call this cognitive bypass, and it presents a real challenge for computing educators.
This workshop introduces coLearn-AI, a platform that structures small-group collaboration so that AI scaffolds student reasoning rather than replacing it. You will first experience the platform as a participant, then learn to author activities for your own courses.
🧠
Explain cognitive bypass and why it matters for computing assessment.
🤝
Describe POGIL-style roles and how they distribute responsibility.
✍️
Design a working activity with feedback prompts and sample responses.
🚀
Plan deployment of coLearn-AI in your own classroom.
🔬
Identify research opportunities and collaborations.
Workshop Schedule
-
8:50
arrive
🔑 Before We Start
Sit with your group (the table number is on your login card). Connect to any Wi-Fi, open colearn-ai.com, and log in with the student account on your card.
-
9:00
15 min
Part 1: Introduction
Cognitive bypass, why struggle and peer talk matter, the design principles, roles and turn-taking, accept-or-revise feedback, and the epistemic trace.
-
9:15
45 min
Part 2: Be a Student
Your table chooses one of two activities, programming (Loops with Turtles) or map reading (How Maps Mislead), and works on it together for 20 minutes. Then a short debrief, and a look at what the instructor view recorded about your groups.
-
10:00
10 min
☕ Break
Back at 10:10 sharp. You return as an author.
-
10:10
75 min
Part 3: Create an Activity
Watch an activity built and revised with an LLM. Then build, test, and save a 2–3 question activity for your own course.
-
11:25
15 min
-
11:40
20 min
What You Need
💻 A Laptop
With a modern browser (Chrome or Firefox work best). No additional software installation is needed.
📶 Any Internet Connection
The workshop runs on colearn-ai.com, so any Wi-Fi works, including the guest network or a phone hotspot. If it is unavailable on the day, we will switch to a backup server on Kenyon's eduroam network.
🔗 Platform Access
You get a login card at your seat, with a student account for Part 2 and a creator account for Part 3. You do not need to create an account. On the Kenyon network you can also look up your login here. In Part 2 you can choose a Python programming activity or a map-reading activity, so no particular background is required.
Audience
Computing educators (especially instructors of introductory programming or data structures courses) who want to incorporate AI into their teaching while preserving meaningful student engagement. No prior experience with AI tools or activity-authoring platforms is required.