Assembling Knowledge: Active Learning in the Age of AI

Useful understanding requires learners to apply ideas, so memorising many digits of pi offers limited evidence. The same problem appears when a student reproduces notes in a familiar examination or learns to phrase a good prompt for an AI tool. Either strategy may produce a correct-looking response. Evidence of readiness comes from handling an unfamiliar problem, weighing evidence, and making a sound judgement.

Active learning gives students time to do that work in class. Replace nineteen minutes of talking and a one-minute activity with a brief frame and a substantial block for exploration. In the furniture exercise, teams receive separate pieces, clues, and instructions. They must coordinate, assemble the object, and reflect on what happened. The relationships become clearer because students have had to find and test them.

The news-headline exercise offers another simple format. Students first write five questions independently before using any device. An LLM produces five more. The class then compares the two sets: What assumptions do they contain? Which questions show curiosity? What has each set missed? Use the comparison to examine assumptions, curiosity, gaps, and question quality.

Prepare the activity carefully. Curate short videos, images, papers, passages, or project fragments. Credit the sources, state the objective, and use a short quiz when accountability is needed. Adapt examples to the learners, discipline, and local industry. Assessment should then ask students to analyse, explain, compare, verify, connect, and reflect. Bloom’s taxonomy or SOLO can help check the cognitive demand. As AI tools change, these actions keep the learner responsible for the thinking.