AI in learning is risky when it hands you answers and saves you the work of thinking. It is like going to the gym and having a robot lift the weights for you. You are missing the point. Science calls this cognitive offloading. It’s when we use tools to reduce mental effort. It can be a good thing, but not when the mental effort is what creates the learning. But it does not have to be that way. A better use is when it slows you down and asks the questions you would otherwise skip.
In this post, I’ll explore three practical questions:
- What can go wrong when AI is used for learning?
- Why is reflection where learning happens?
- How can AI support reflection in practice?
The most impactful learning comes from experience. Well-designed learning activities are those that create meaningful experiences. But the experience still needs to become learning, and that is where reflection plays a crucial role. In Kolb’s experiential learning cycle, reflection is the step that turns experience into learning. This also aligns with what learning science calls generative learning. The learner has to do the sense-making work. The challenge is that reflection is also the step that is easiest to skip or to do poorly.
Three rules of thumb:
- Don’t let AI do the thinking for you.
- Make reflection a distinct step in the flow.
- Keep it coaching-style: ask, listen, then follow up.
This is the approach we use with Reflect with AI at TalentMiles. The AI engages participants in learning programmes in a coaching dialogue, prompting them to slow down and reflect on the experience they just had. Instead of having an AI write answers for you, the AI asks you questions to help you think through what you have learned. Because the AI adapts to the participant’s responses, it can ask follow-up questions that help them dig deeper in a way that a static form cannot.
The experience for the participant goes something like this. The participant finds a task in the TalentMiles app. They read the task, which gives them an activity to do. Once they have done the activity, they return to the app. Before they can submit their answers, they first have to engage in a dialogue with an AI that asks coaching questions and encourages them to pause and think about the activity they just did. For example, it might ask: “What felt harder than you expected, and what might be going on there?” and then follow up based on the participant’s responses. After this dialogue, the participant can proceed to submit their answers to the task.
The AI doesn’t replace the learning coach. The coach still shapes the reflection for each task by setting the focus and tone, and the AI carries that out as a coaching dialogue in the moment. The participant’s answers for the task are still submitted to the learning coach as usual.
This might strike some as just busywork, but the speed bump between doing the activity and submitting the task helps protect the moment of reflection. In today’s busy world, it is too easy to rush through and lose out on the learning available. Now, to be fair, this system cannot force the participant to take full advantage of the opportunity provided. It is still incumbent on the participant to choose to reflect. The AI can only prompt and encourage.
Reflect with AI scales well. The AI can provide every participant with a coaching dialogue whenever it fits their schedule, no matter how many participants there are in the programme. This reflection step can improve learning within tasks and increase the overall impact of the programme. For the participant, it provides a calm bridge between “doing the task” and “answering the questions” that lets them take a moment to make sense of what they just experienced. The secret to using AI in learning and development is that better questions beat more answers.
