AI in Education, Critical Thinking, and the Future of Teaching with Alec Couros

On today's episode of Roll Call, Brian and Josh welcome Dr. Alec Couros, professor of educational technology and media and director of the Centre for Teaching and Learning at the University of Regina, for a deep and timely conversation on the role of AI in education.

The conversation explores one of the biggest challenges facing schools today: how to integrate AI while maintaining strong pedagogy and human connection. Alec shares insights on cognitive offloading, critical thinking, misinformation, and why writing still matters more than ever in a world where AI can generate almost anything instantly.

They also unpack the risks of over-prioritizing efficiency in education, the future of personalized learning, and why educators need to be intentional about how these tools are used — not just what they can do, but what they should do.

Show notes:

00:00 — Introduction to Alec Couros

01:11 — Alec's background and how he found his path into education

04:16 — Balancing AI integration with strong pedagogy

05:27 — "Pedagogy first, technology second"

07:18 — Is critical thinking changing in the age of AI?

08:56 — AI grading, black-box processes, and bias in assessment

14:23 — Productive struggle, learning, and cognitive offloading

17:42 — When students should be introduced to AI tools

19:18 — Why writing is thinking

23:18 — Misinformation, deepfakes, and teaching digital discernment

28:22 — What keeps Alec up at night about AI in education

29:35 — Efficiency, teacher perception, and controlling the narrative

31:58 — Personalized learning, isolation, and protecting human connection

33:25 — AI tutors, non-certified guides, and the future of school models

36:31 — Curriculum upheaval and what is still worth learning

39:25 — The pace of AI development and whether educators need to keep up

44:52 — The future of personalized learning in the classroom

51:58 — Lightning round

55:11 — Final reflections on alignment, values, and what AI ought not do