Adaptive Learning

Every learning app you have ever used asks you exactly one question about yourself.

It wants to know whether you got the last thing right or wrong. That is it. From that single signal it decides whether to make the next question easier or harder, whether to show you more of the same or move you on. The whole industry calls this personalised learning, but it is personalised in the way a shoe shop is personalised when it asks what size you wear and nothing else. You walk in as a complete person with a lifetime of experience and preferences and moods and the thing you were thinking about on the way there, and the system reduces all of it to a single number on a sliding scale between too easy and too hard. Then it congratulates itself on knowing you.

Real learning does not live on a single axis. You arrive at a topic with prior knowledge that might be partial or flat wrong, with motivation that changes depending on what else is going on in your life, with a preferred way of making sense of things that has nothing to do with how quickly you can answer multiple choice questions. A system that adjusts only difficulty is adapting to your last click, not to you. This is the gap that most adaptive technology still has not closed, and it is why people who use these apps often feel they are being measured but not understood. When was the last time a piece of technology seemed to actually know what you needed rather than just what you got wrong?

Last week we asked what a single tick on a screen really tells us. This week we look at the polite phrase personalised learning and at how narrow most of it really is.