Redesigning Learning for the AI Age

Outline

1. Knowledge is not isolated Every concept needs links to definitions, problems, claims, formulas, and experience.
2. The learning sequence should change Acquisition, understanding, expansion, correction, and application can all be redesigned.
3. AI should act as structured practice It should expose gaps, errors, and transfer paths instead of replacing the work.

Notes

This is a draft for a long-form video about redesigning the whole learning sequence once AI enters the loop.

I prefer treating AI as structured practice: it can expose blind spots, organize concepts, create exercises, and build counterexamples, but it cannot replace transfer.

Slides

1. More information does not guarantee understanding AI makes acquisition cheap. The scarce parts are judgment, connection, correction, and transfer.
2. Learning needs order Get accurate information, understand where it sits, expand the model, correct errors, then use it in a real situation.
3. Make AI a structured practice partner Ask it to question, find gaps, rephrase, build examples, and reverse-test your understanding.