Tutoring, lectures, feedback on copies, grading: from the student's point of view, the academic uses of AI can be reduced to four functions, each of which hides a part of the pedagogical relationship.
At Macquarie University in Australia, a conversational agent called "Virtual Peer" has been introduced in two psychology teaching units: part of the content previously delivered in tutorials by videoconference now goes through the chatbot and online activities. Other Australian institutions — Sydney, Wollongong, La Trobe, Newcastle — are experimenting with similar devices for feedback on copies, tutoring or the course itself.
Thomas Corbin, author of this analysis published on September 8, proposes to reduce these uses to four functions from the student's point of view: tutoring, teaching, commenting on a piece of work, and grading. The benefit is real and easy to describe—feedback available at any time, valuable linguistic support for non-English-speaking students, a constancy that human fatigue never guarantees.
The loss is more difficult to formulate, and this is where the argument lies: a system can produce a clear and useful comment without acknowledging the student who wrote it, that is, without situating this work in a trajectory, or adjusting its teaching to a particular person. The article invites universities to name precisely what they delegate, rather than arbitrating en bloc for or against the tool.
Read the full article on the original source
Go to the source →