Teacher Feedback at Scale? A Promise of Boring AI for Digital Learning
In my twenties, I was involved in a project teaching digital skills to community groups and unemployed school-leavers. The work was gratifying, but most of my time was spent working one-on-one with programme participants, providing them with individual feedback. As you can imagine, my interventions were limited to the people in the room.
Now think about a large undergrad class, such as Marketing or Business Studies, where a professor might have upwards of 700 students. Each student will need to complete course activities and assignments. This professor will struggle to provide individual feedback, focusing primarily on marking submissions and giving general feedback (unless a student reaches out with an issue). It’s worth noting that feedback is substantially more effective when received close to the time of the activity.
But could there be a better way to do this? One that leverages both automation and natural language processing. I call this Boring AI, which uses AI to complete time-consuming or mundane tasks that provide real-world benefits.
Take this scenario, for example, where a teacher creates a custom GPT using ChatGPT Plus. If a teacher writes a prompt for this GPT based on their approach to providing feedback, along with examples of good, okay and bad assignments and their feedback is given, could this create a twin AI tutor version of themselves, specifically geared towards providing insightful feedback on student work at scale? The teacher could then focus on individual interventions for struggling students or personalised coaching to students benefiting from additional extension activities.
And this now responds to two issues — the teacher who doesn’t have time to provide feedback to each student in their large classes and the students who benefit significantly from receiving immediate feedback on completed project work.
Follow me on X/Twitter for bit-size comments and (not guaranteed) lols: twitter.com/niall_mcnulty_
