Learning Paths
When the answer is no longer enough
If answers are available everywhere, what exactly is the teacher's role? But there is another side to that question. As teachers, answers have always been useful to us too. I give you a problem. You solve it. I look at your answer. And from that, I make some judgement about whether you understood what I taught. That worked reasonably well when producing the answer itself required some effort. Today, I'm not so sure. Suppose I give a student a physics problem. The student comes back with the correct answer. Perfect working. Correct formula. Correct units. Beautiful explanation. What exactly have I learned about that student? Maybe she solved the whole thing herself. Maybe she struggled with it, made two mistakes, finally understood something and then arrived at the answer. That would be fantastic. Or... maybe she asked AI. Or maybe she attempted half of it, asked AI for help, understood the explanation and completed the rest. That might be perfectly fine too. Or maybe she copied the entire solution without understanding any of it. All three situations can produce almost exactly the same final answer. And that's the problem. The answer has become weaker evidence of learning. I don't mean answers are useless. Of course the final answer matters. If you're designing a bridge, writing software or calculating somebody's medication, eventually getting the answer right is extremely important. But when I'm trying to understand whether somebody is learning , I need to see more than the final output. I need to see some part of the journey. So perhaps instead of only asking: "What answer did you get?" I also start asking: "How did you get there?" "What did you try first?" "Where did you get stuck?" "Why did you choose this approach?" "What did you initially think was happening?" "What changed your mind?" And one of my favourites: "What would happen if I changed one thing in the problem?" Because that's often where understanding becomes visible. A learner might have a perfectly polished answer. But change one condition... and suddenly you discover whether the idea has actually been understood. Or ask: "Can you explain this to me without using the formula?" "Can you give me another example?" "Can you compare two approaches?" "Which solution do you think is better, and why?" Now I'm not simply examining the output. I'm looking at the learner's thinking. And I think this matters enormously in the age of AI. Because we have become very good at judging finished work. Assignments. Reports. Code. Presentations. Essays. Solutions. But AI is becoming extremely good at producing finished work too. So perhaps education has to become better at noticing what happened before the finished work appeared. The attempts. The questions. The wrong turns. The reasoning. The feedback. The correction. The second attempt. Those things can sometimes tell me much more about learning than the polished answer at the end. And I want to be careful here. I'm not suggesting that we start monitoring every single thing a learner does. That's not the point. And I'm certainly not suggesting we make students prove that they never used AI. I actually think AI can be part of the learning process. The important question is: What intellectual work did the learner still do? Did AI help her understand something she was struggling with? Great. Did it challenge her assumption? Great. Did it give her feedback on an attempt? Wonderful. Did it allow her to skip the thinking entirely and hand over a finished answer? That's different. And I think teachers will increasingly have to design learning experiences where that distinction becomes visible. Maybe assessment itself has to change. Instead of one large finished submission... perhaps there are smaller checkpoints.
