Teaching to the test
On cram-studying, what people are willing to believe, and why turning to ChatGPT for all the answer is the wrong idea
The obvious reason to object to the over-use of artificial intelligence tools in the classroom is that it looks almost self-evidently like an abandonment of duty. Teachers, especially in advanced professorial positions, are there in no small part because they are expected to have superior content knowledge to share with their students.
■ But good educational experiences depend upon more than just content knowledge alone. Human learning depends heavily on motivation: What meta-questions are asked more than “Why do we have to learn this?” and “When will I ever need to use this?”. It’s not just a question of shirking: Most people need to feel some kind of self-interested motivation in order to learn successfully.
■ In testing-dominated environments, the test becomes the motivation: “I have to cram for this test so I can pass the class”. Material “learned” that way rarely sticks around long enough to migrate into long-term memory. No matter who the learners are, they benefit from teachers who understand and care about their motivationss.
■ No computer can truly “understand” motivation. They can copy the work documented by human beings on the subject, but there’s no understanding first-hand, the way any halfway competent human can gather it just from walking into the room. The elusive qualities of “energy” in a room aren’t magic, but they aren’t digital, either.
■ Fundamentally, this is why any attempt to teach by substituting technology for good human judgment is ultimately going to ring hollow.
■ Sure, technological tools can be used as aids, but an attentive human teacher will always be the overall superior tool for teaching other people.
■ There just isn’t a way to effectively substitute for the good common perceptive sense of a teacher in human motivational feedback, and short of giving computers real sentience inside real bodies, that advantage will always belong to people.