TLC-FMG AI Tools
Generative AI has changed what we can reasonably ask of students, and most of us are still working out what that means for our own courses. At the TLC-FMG, we have been building a small set of open source tools to support colleagues with integrating GenAI responsibly into their course, and to make their course resistant against unethical GenAI usage. Now they all live in one place at tlcfmg.github.io. On this page you will find a short exposition on the different tools available.
AI Exposure Scan
AI Exposure Scan helps teachers reflect on the assessment structure of their course. Select your forms of assessment and the tool calculates a dynamic AI exposure score, giving an indication of how susceptible these assessments are to generative AI use.
Learning Goals Reality Check
A worksheet for testing whether a specific assessment still meets its learning goals in an AI-rich world, and refining it until it does.
Saurabh demonstrates Canvasfeed
AI Detective
Are you able to distinguish human academic abstracts from AI-generated ones? Are you sure? We made a quiz in which two abstracts will be shown: one AI-generated, and one from a human. What will be your record?
Canvasfeed
Turns a Canvas discussion board into a structured summary and a personalised follow-up question for each student, ready to use in class. Note: Canvasfeed needs some technical know-how and is thus not for the faint of heart.
Surveychat
Lets students run surveys and randomised experiments, embedded in platforms such as Qualtrics, with LLM-based conversational agents without building a web application.
All tools are free and open source. Have a look, try them out in your class, and let us know what would make future tools more useful for teachers.