On Episode 584, I welcome Danny Liu, Professor of Educational Technologies at the University of Sydney, to Teaching in Higher Ed. We explore what it means to trust students in the age of AI and how traditional assessment metaphors like traffic lights often fall short in guiding student use of artificial intelligence. Danny Liu introduces his “menu” framework, inspired by his own experience on a Disney cruise, to help educators guide students through the many choices AI presents—emphasizing agency, healthy decision-making, and care.

Together, we discuss the dangers of corroded trust between students and professors when using surveillance-driven assessment practices, and how embracing both open and secure forms of assessment can better prepare learners for real-world challenges. Danny and I reflect on practical strategies for fostering trust, student agency, and authenticity in assignments—moving away from attempts to fully control students’ use of AI and instead equipping them for ethical engagement.

Resources from the episode:

• Menus, not traffic lights: A different way to think about AI and assessments, by Danny Liu
• Talk is cheap: why structural assessment changes are needed for a time of GenAI, by Thomas Corbin,Phillip Dawson, &Danny Liu
• What to do about assessments if we can’t out-design or out-run AI? by Danny Liu and Adam Bridgeman
• Course: Welcome to AI for Educators from the University of Sydney
• Whitepaper: Generative AI in Higher Education: Current Practices and Ways Forward, by Danny Y.T. Liu, Simon Bates
• Five myths about interactive oral assessments and how to get started, by Eszter Kalman, Benjamin Miller and Danny Liu
• Interactive Oral Assessment in practice, by Leanne Stevenson, Benjamin Miller and Clara Sitbon
• ‘Tell me what you learned’: oral assessments and assurance of learning in the age of generative AI, by Meraiah Foley, Ju Li Ng and Vanessa Loh
• Interactive Oral Assessments: A New but Old Approach to Assessment Design from the University of South Australia
• Interactive oral assessments from the University of Melbourne
• Long live RSS Feeds
• New AI RSS Feed
• New AI RSS Page
• Broken: How Our Social Systems are Failing Us and How We Can Fix Them by Paul LeBlanc

Episode topics:

• Trust and the Student-Professor Relationship in the Age of AI
• The Disney Cruise Menu: A Metaphor for AI Choices in Learning
• Why Traffic Lights Don’t Work for AI Assessment
• Navigating Student Agency Amidst Artificial Intelligence
• The Pitfalls of Surveillance and Over-Policing Assessments
• Open versus Secure Assessments: A Balanced Approach
• Embracing Variability: Personalized Learning Paths with AI
• Modeling Authentic AI Use as Educators
• Rethinking Feedback and Care in the AI-Influenced Classroom
• Assessment Design for Real-World Learning and Integrity
• Building Student Confidence and Curiosity Around AI Tools

Discussion questions:

1. Danny Liu talks about replacing the “traffic light” model for AI assessment with a “menu” approach. What are the core differences between these models, and why might a menu be more effective in today’s educational landscape?
2. Both Bonni and Danny emphasize student agency and trust. In your view, what strategies can educators use to foster trust with students, especially in the age of AI?
3. Danny notes that assessment scales and rules around AI usage can fall short because students may simply ignore or bypass them. What realities of student behavior and learning does this expose, and how might educators respond constructively?
4. Danny suggests that secure, in-person assessments can validate learning outcomes, while open assessments may embrace more autonomy. How might a balance between these two approaches look in your discipline or context?
5. What’s your response to the idea that varied “diets” of assessment—moving beyond just exams or just essays—can better support authentic learning in the era of AI?
6. The conversation highlights using AI as a feedback or preparatory tool for students, but not as an automatic grading tool for summative assessments. What are the potential benefits or drawbacks of this stance?

Related Episodes + Resources

All the AI-related episodes that have aired on Teaching in Higher Ed may be found in one spot: AI-related episodes. You can also find many related resources on the Teaching in Higher Ed main AI page.

Quotable Words

Danny Liu stresses on Episode 584:

Our students are presented with this massive array of things they could choose from. They may not know the right things to choose or the best things to choose. And our role as educators is to guide them in trying to find the most healthy options from the menu to choose from.

Next Week’s Episode

On the upcoming episode of Teaching in Higher Ed, Bryan Dewsbury returns to the show. This time, he points us toward socially just teaching across disciplines.

Support

The money gathered via the TiHE virtual 'tip jar' helps to defray some of the costs of producing the podcast.

Read

My book: The Productive Online and Offline Professor: A Practical Guide, provides approaches to help you turn your intentions into action. I also write an advice column for EdSurge: Toward Better Teaching: Office Hours With Bonni Stachowiak

Listen

Subscribe to the Teaching in Higher Ed podcast and listen on Apple Podcasts, Google Podcasts, Overcast, Stitcher, TuneIn, or Spotify.

Share

Update: If you enjoy reading these weekly updates and would like to share them with a friend, they can sign up on the Teaching in Higher Ed updates subscribe page.

Disclosures

Affiliate income disclosure: Books that are recommended on the podcast link to the Teaching in Higher Ed bookstore on Bookshop.org. All affiliate income gets donated to the LibroMobile Arts Cooperative (LMAC), established in 2016 by Sara Rafael Garcia.”

Notice: Portions of these weekly updates are produced using CastMagic.io, which uses AI to produce a draft of the transcript, identify key quotes, highlight themes, etc.