My hot take: we shouldn’t use essays as a metric of learning in schools.
We can keep them as products, as they are still useful, but they are horrible at capturing what the student actually learned.
MIT’s research finds that if AI is used to lead critical thinking in essay writing, students don’t remember what they wrote in the essay (an indication that students are cognitively offloading too much and in ways harmful to their learning).
I don’t blame the AI for this. I blame the way we’re currently measuring learning, which is often product-based.
Let’s remember what the essay is: an artifact of someone’s thought process and critical thinking on a topic. They capture thinking that has happened, has occurred, has been refined and defined. It’s past tense.
AI now can recreate that artifact. In fact, it can create a lot of artifacts. So, if we want to assess someone’s learning process or critical thinking on a topic, we need to find a dynamic vehicle that follows the cognitive process along the learning path.
Schools and tools are starting to test this with some promising results. Schoolai.com has Dot, a tool that interacts with students like an assistant teacher, who can then share back with the teacher where the student is. It evens alerts teachers to the emotional safety of the student, alerting teachers on things like bullying, abuse, or neglect. Khan Academy is doing the same thing with Khanmigo.
Higher Education institutes are doing something similar: University of Michigan’s Center for Academic Innovation supports instructors to “[e]mphasize the learning process over the product”. The Online Learning Consortium recommends using viva-style check-ins for code, essays, and capstone projects
That’s what we get to do as educators: find new ways to authentically capture the learning process and critical thinking. What is not our job is to figure out how to maintain the life of the essay.