Technology Finally Kills the Thayer Method of Teaching: Troubling Times for Educators

(The opinions expressed in this article are those of the author and do not necessarily reflect those of Al-Fanar Media).

Generative AI (GenAI), most notably in the form of large language models (LLMs), has had a dramatic impact on education, and, in general, educators and educational institutions are not responding well. It appears that many are not responding at all or making only patchwork responses.

This tepid reaction may be due to experiences with prior technologies, such as calculators, computers, and search engines, that initially seemed to upend traditional education but ultimately led to little change in many classroom interaction methods. Even the telecomputing technology brought on by Covid-19 had a limited impact in accelerating the delivery of courses online.

Some of this lack of change was due to the resilience of some instructional methods, such as the Thayer method of teaching.

Table of Contents

A Student-Centred Approach

The Thayer method is a student-centred teaching approach that shifts much of the responsibility for learning factual information to the student. In this method, students are expected to study new material before the class meeting, bringing questions and being prepared for evaluation. Class time is reserved for collaborative problem-solving, active learning, discussion led by both teachers and students, and evaluation. One can see elements of the Thayer method in approaches of problem-based learning, the flipped classroom, and maker spaces.

The Thayer method was developed at the U.S. Military Academy at West Point, where I earned—and “earned” is the correct word!—my undergraduate degree. I can still recall walking into, for example, a math class, sitting at my seat, and then having the professor say, “Take boards!”. We students would then get out of our seats, go to the blackboards, individually, and work on a math problem. After a given amount of time, the professor would select one student to present their solution to the class, walking through their methodology for solving the problem. No lecture. No slides. No video. Active learning.

Similar occurrences happened in chemistry, computer science, history, engineering, law, and philosophy courses. Not a lot of lectures. A lot of discussions. A lot of evaluation. A lot of explaining. A lot of conversation.

Given this no-lecture approach, the Thayer method was very resilient to earlier technologies. The calculator, computer, and search engines had little effect on the Thayer method, other than making it easier (in theory) for students to prepare for class, as accessing information was now more convenient.

Removing a Key Human Element

Large language models, however, are different.

What is the difference between LLMs and earlier technologies?

Dialogue.

LLMs are a subset of generative AI that are trained on vast amounts of data which they can draw on to respond to human queries in a conversational manner. They have moved information access from documents to dialogue and from retrieval to reasoning. This conversational aspect differs from earlier technologies, which merely made information easier to access, and is especially disruptive to even resilient teaching methods, such as the Thayer method.

Conversation has been a uniquely human-to-human form of information exchange for tens of thousands of years. We humans are wired for this type of information exchange. With no technological capability for replicating this unique human form of information exchange, teaching methods like the Thayer method weathered successive technological introductions. None could replace the dialogue that was uniquely human.

Not anymore. LLMs now possess these communication capabilities, and they are quite good at it! They can question responses, evaluate statements, review formulas, and engage in discussion as a collaborative partner. In short, they can do almost everything that made the Thayer method so resilient in the past.

Questions for Educators

Therefore, we educators must ask ourselves: “What does educating mean when technologies no longer just retrieve but infer, collaborate, and create with humans in a conversation?”

Answering this question is the critical first step in understanding how we can live with, manage, and harness the power of AI for the educational benefit of students, enabling them to face the many personal and planetary challenges ahead.

Bernard Jansen is principal scientist at Qatar Computing Research Institute (QCRI), a national research institute that operates under the umbrella of Hamad bin Khalifa University (HBKU).

This piece has been submitted by HBKU’s Communications Directorate on behalf of its author. The thoughts and views expressed are the author’s own and do not necessarily reflect an official university stance.