Alan argues that the true value of AI in education lies not in increasing the volume of information consumed, but in increasing the complexity of the knowledge a learner can master.

By using AI as an “intelligence amplifier”—specifically through high-fidelity PDF parsing, large context windows, and spatial organization on digital whiteboards—learners can bypass the “shallow” popular summaries of a field and engage directly with dense primary sources (e.g., PhD-level textbooks).

In an AI-saturated world, efficiency gains should be reinvested into harder material, not more material.

Surprising or Counter-Intuitive Points

• Summaries are Bad for Math: While summaries provide a “compass” for humanities, they are often useless for mathematics, where understanding is derived from the process of proof and definition rather than the conclusion.
• AI Prevents Thinking Outsourcing: Counter-intuitively, using AI to provide “hints” during difficult problems keeps a learner persistent and motivated, whereas struggling in isolation often leads to giving up (the ultimate failure of thinking).
• The Primary Source Imperative: AI makes it easier to avoid “middleman” summaries and misinformation by lowering the barrier to entry for dense, authoritative academic texts.