What Should University Education Do in the Age of AI?

I encourage my students to use AI. We already use AI in many aspects of our lives, and the classroom should be no different. But the role of university education has changed in the age of AI. It is no longer about transmitting knowledge — it is about helping each student learn to build their own narrative.

Why I Am Rewriting the Textbook

I teach genetics in the spring semester and programming in the fall. Both are sophomore-level courses. For genetics, I never adopted a standard textbook. Instead, I assembled my own materials — key papers, reviews, woven into slide decks — and taught from those for six years. Now, during my sabbatical, I am writing a textbook to reshape how I teach.

I am skeptical of slide-driven lectures and rote memorization. Biology is a rapidly evolving field. In genetics alone, paradigms have shifted multiple times over the past decade as massive datasets emerged. Rather than memorizing knowledge that changes so quickly, it is far more important to understand how perspectives on phenomena are formed and how they evolve. So in writing this textbook, I focused on the context and evolution of concepts rather than rigid definitions. Starting next year, I plan to eliminate slides entirely and lead discussions around a few key figures.

What Remains for Humans

As ChatGPT became widespread, I often asked myself: what can humans do better than AI? There is no clear-cut answer, but one thing is certain — humans have curiosity and desire about specific topics. AI possesses vast knowledge, but it cannot generate the question "I wonder about this" on its own. University education should therefore be a time for students to discover their own desires and learn to express them through speech and writing. This is precisely where university courses can differentiate themselves in the age of AI.

An Experiment in the Programming Course

In my fall programming course, students spend the first three weeks learning basic coding while simultaneously being asked to find a paper whose questions they want to answer. They then explore the publicly available data from that paper — not raw data, but summary data from supplementary materials.

But before looking at any data, I ask them to draw by hand what they want to analyze and visualize. I hand out paper in class and have them sketch. They explain their drawings to their peers. Over the remaining weeks, they analyze the data, write code, compose text, and complete a portfolio. The finished two- to three-page portfolios are exchanged among students for peer review.

This kind of coursework used to be difficult, but since 2024, with ChatGPT available, even sophomores can accelerate their code learning and portfolio writing substantially. So I explain how to use AI for writing code, reading papers, and composing text.

This type of teaching is impossible in large lectures. Forty students is my maximum. During the programming course, I set up a Slack workspace where students can ask questions 24/7. Over a semester, roughly 8,000 messages are exchanged. I also invite students to a team Notion page to document what they learn. This lets me observe the direction and extent of each student's intellectual growth over the course of the semester.

Why Come to University?

This is a question I have openly posed to my students every semester for the past six years. My answer is this: university should be a place where each student builds their own story. Many people point to the "individualism" of Gen Z as a weakness, but I see it as a significant strength. The students I have taught have a strong desire to create their own narratives. When given the task of writing their own questions and interpretations through programming portfolios, they all performed capably — technical difficulties aside. At conferences too, the poster presentation attitudes and participation levels of today's students far exceed those of previous generations.

What Is Needed Is Freedom, Not Regulations

What students need to learn at university includes structured knowledge, of course, but more importantly, the ability to translate that knowledge into their own language and build their own stories. Students will each choose their own futures — they will not all study the same topics as I do or pursue the same career. At university, the starting point of that journey, they need to form their own thinking.

For this kind of education to work, what is needed is not more regulations but more freedom. What professors need is the time and freedom to study, research, and teach. During my sabbatical, I wrote several textbooks and made them all freely available (chaek.org). In the age of AI, university education should be a place where each student builds their own narrative. What educators need for this is not new systems, but freedom and time. Give each instructor maximum individuality and freedom, and students will find their way to building their own stories.