Teaching in the Time of ChatGPT

2026-04-15 3:46 PM

#llm #teaching #comment

Original link

This Ars Technica article has been breezing around the Internet for a couple days and I finally had some time to sit and read it this morning. It's full of very quotable nuggets that, more than anything, communicate the frustration teachers all over the world are feeling.

But since the appearance of ChatGPT, the instructor’s job isn’t just to teach the subject and frantically attempt to keep every student’s plate spinning. Increasingly, it’s to moonlight as a detective and prosecutor because students without the motivation to do the work don’t have to skip it anymore.

This kind of sums up the essay - there's an enhanced air of suspicion about every piece of work that comes in from students these days. My situation is quite different - I'm face to face with a much smaller number of students. However, using standards based grading, I still look at a lot of student work. I have to see evidence of understanding - can they do the thing described by the current learning goal?

The hard work of lifting weights is the point because that yields physical results. A popular analogy is that using an LLM to write your essay is like driving a forklift into the weight room. Weights get lifted, sure, but nothing is accomplished.

I talk with my students about "productive struggle" all the time. My practice pages don't have points attached in the gradebook because the entire purpose is for me to see authentic work, right or wrong. If it's right, I can confirm their work and give suggestions on clarifying processes or ask follow up questions to help them connect concepts. If it's wrong, I can help them fix those mistakes. If it isn't their work - copied or LLMd or Googled - then I cannot do my job. This is a real mindset change, especially for 10th graders. I remind them that I cannot force anything, but they're still responsible for the skill come assessments.

Many instructors are trying to adapt to this crisis by going back to the only evaluation tools that are pretty much LLM-proof—tests like oral exams or handwritten work created under supervision in the classroom.

None of these solutions are available to instructors of asynchronous online classes. That sucks, since the availability of those classes is important.

This is the hardest part of the article. I don't have online classes and my tests are handwritten, in class, in front of me. I know that the assessment is their best representation of understanding and I can point back to their evidence of learning (ie, the practice they've done before the test) as a way to promote reflection. "What did you do to prepare?" or, "How does this result match your prep work?" are common in my discussions after exams. It brings to the forefront the danger of offloading small tasks because all things come around again.

I don't think education is in the middle of an AI revolution and I don't think I need to radically change what I'm doing. These fads have come and gone and will continue to do so. I think the bigger question is how we'll all weather this current trend and, more importantly, how we'll get students to reflect critically in their participation.