Teaching in a ChatGPT World

By Kathryn P. Haydon

Educators, Grow Your Creativity

What if ChatGPT can prompt educators to make it impossible to cheat AND to improve teaching and learning?

Instant Cheating with ChatGPT

Over 30% of college students admit to having used the new artificial intelligence tool ChatGPT, launched in late November 2022, on homework, tests, or to write essays. One study found that 89% of surveyed students had used it for homework. Students themselves call this cheating and, at the very least, the practice introduces ethical questions.

After rampant cheating at Stanford and Harvard during final exams in late 2022, the maker of ChatGPT introduced an “AI Text Classifier” that educators can use to detect ChatGPT-generated work. However, this tool isn’t always needed. Much of what ChatGPT currently writes sounds eloquent but is factually incorrect.

Why ChatGPT is Not Reliable

ChatGPT uses a large language model but doesn’t know about truth or fact. It can’t remember or reason. It is appropriately termed artificial intelligence.

That said, AI is said to improve over time with use. As it ‘learns,’ its outputs should become more accurate. With technology like this, it’s not hard to see the role of the teacher devolving into that of a cheat-detection officer—a losing proposition because students and technology will always find new workarounds.

Why We Need to Train Critical, Creative Thinkers

Even before ChatGPT, student cheating was rampant. Several studies put that number at 86% of college students. We need to make a change in our approach to education.

When it is so easy to cheat, school work is not requiring students to exercise high levels of thinking. When it is so easy to cheat, we are not effectively testing students on their ability to think critically and creatively.

ChatGPT itself presents us with a stark reason why we need critical thinkers: its output cannot be trusted. If we blindly trust the output of ChatGPT (or even “Hey Siri”), we are accepting information very uncritically, like reading a study without footnotes. The sources are dubious at best. At least with Google we have to vet websites as we scroll through the results.

Education for Humans

Another problem with technology is that it tempts us to be more like it. We think we need to become more like machines: fast, efficient, mechanistic. But in reality we need to do the opposite.

Nearly six years ago, I interviewed Wired Magazine founder and Silicon Valley tech guru Kevin Kelly about learning. He said that as technology proliferates, our job is to “be more human.” We need to leverage our unrepeatable human characteristics like critical and creative thinking, compassion, love.

Specifically, Kelly told me that he envisions, “. . . teaching restored in some sense to a high-status position. . . not necessarily teaching information but teaching all the other things—like asking questions—that young minds need to know.”

The great news is that we have a model for this and it’s already being used in many classrooms around the world. Imagine Socratic circles, oral exams, and discussion. Imagine, in our increasingly impersonal tech-world, students in small groups with teachers who are well-versed across disciplines and trained to lead discussions that provoke deep thinking and questioning. Think of the flipped classroom model, wherein students watch lectures or consume readings at home so they can use precious class time to engage in deeper discussion and in-depth projects.

When we embrace our most valuable human characteristics, we don’t need to compete with tech. We don’t need to accept roles as tech overseers. Instead, we can create educational settings that develop the essential human characteristics of critical, creative thinking and original thought. After all, so-called intelligences like ChatGPT can only spit out what we put in.

Glad you asked!

Yes!! We do train teachers how to teach high-level thinking that can’t be Googled or “ChatCPT’d.” Click here to find out which PD program is best for your faculty!