One of the best experiences of academic growth that I had was because  of my high school English teacher who took time to write specific comments on  my essays. Back in the 2000’s when typing a paper required a human, I wrote those essays myself in Word on a boxy desktop computer. My teacher, of course, invested large amounts of his time to read our essays and provide thoughtful comments.  

Personalized feedback hasn’t gotten cheaper or easier with time. As Russ Roberts asks in this episode, “How do we scale grading and feedback?” Until recently, it’s been a bottleneck, constrained by the availability of skilled human time. We can’t (yet) completely automate empathy, discernment, or pedagogical intuition.

For now, the question is still what kind of feedback teachers can give that really benefits students. Daisy Christodoulou, the guest on this episode, offers a sobering critique of how educators tend to give feedback in education. One of her points is that much of the written feedback teachers give is vague and doesn’t actually help students improve. She shares an example from Dylan William: a middle school student was told he needed to “make their scientific inquiries more systematic.” When asked what he would do differently next time, the student replied, “I don’t know. If I’d known how to be more systematic, I would have been so the first time.” The teacher knows what a more “systematic” essay would look like, but the student has (presumably) not done specific practice exercises that would help them achieve mastery. 

Christodoulou articulates that students who aren’t doing well often don’t know how to get better, and generic feedback like “try again” or “be clearer” offers no path forward. I loved her metaphor comparing writing to marathon training. You don’t train for a marathon by running one every day. You build up to it with a mix of shorter runs, strength training, stretching—activities that don’t even look like running but are essential to running well. Similarly, becoming a better writer doesn’t always mean writing another full essay. It might mean building vocabulary, practicing sentence construction, or doing targeted inference work.

Christodoulou emphasizes the need for teachers to think in terms of models of progression—to identify the small, specific steps that move a student from where they are toward mastery. One example she gives: a student told to “infer more insightfully” might not need another essay assignment, but instead a focused set of lessons on vocabulary, prefixes, and suffixes. That’s what meaningful feedback looks like—it leads to action.

Christodoulou also turns to the question many of us are now grappling with: can AI help scale meaningful feedback? Can it grade essays in a way that’s actually useful? As of this recording in early 2025, the answer is: not quite. There are still issues with accuracy, consistency, and what AI models tend to “hallucinate.” Some of those problems may eventually be solved—but even now, Christodoulou and her team are experimenting with hybrid models where teachers give audio feedback and AI transcribes and organizes it. This kind of collaboration, where the teacher remains the source of insight and the AI handles the labor-intensive part, might be one path forward.

At the same time, there’s a growing tension in classrooms: students increasingly want to skip the hard work of writing altogether and turn to AI-generated answers. So, we’re trying to scale authentic feedback just as students are even more tempted to shortcut the learning process because of how easy it is. That challenge of how to preserve the value of thinking and writing in the age of instant text generation is one we’ll be wrestling with for a while. As a teacher, Christodoulou has encouraged me to think beyond “let me show you again” for my students. I will try to break down smaller exercises that will help them achieve mastery of new advanced skills.