How to Learn Anything Faster With AI

Last semester a student came to my office hours confident.

He'd been following AI for months. Newsletters, YouTube, podcasts. He knew every model release, every funding round, every viral demo. He could name the companies, the founders, the benchmarks.

I asked him one question: explain how attention works in a transformer.

He froze.

Not because he was unprepared. Because he'd been consuming AI content for six months without actually learning any of it. The information had passed through him and left nothing behind.

I see this constantly. And if I'm honest, I catch myself doing it too. Opening a newsletter, reading it end to end, feeling informed, and three days later having nothing to show for it except a vague memory that something interesting happened.

AI has made this worse, not better. It's never been easier to get a summary of anything. Which means it's never been easier to feel like you understand something without doing the cognitive work that actually creates understanding.

This guide is the system that fixes it.

Here's the counterintuitive part.

Active recall produces 50-150% better long-term retention than passive methods. Re-reading and highlighting are rated low utility by cognitive science research - they don't create the processing that leads to retention.

Reading a summary Claude wrote doesn't count as active recall. Watching a YouTube explanation doesn't either. Skimming a newsletter and nodding along definitely doesn't.

Real learning happens when your brain has to retrieve information from memory - not recognize it on a page, not skim past it in a feed, but actually pull it out of nothing when someone asks you a direct question.

AI tools make the passive version so frictionless that the active version gets skipped entirely. You get the summary. You get the overview. You get the gist. And you mistake that gist for knowledge.

The system below uses AI differently. Every step forces retrieval, explanation, or application. Zero passive consumption.

The rule behind the whole system: If you can't close every tab and explain something out loud from memory, you haven't learned it. You've encountered it.

Before touching a single article or paper, spend three minutes building the skeleton.

New information needs structure to attach to. Without a map, everything feels equally important, nothing sticks, and you finish reading with a vague sense of understanding that evaporates within 48 hours.

Become a paying subscriber to get access to this post and lots of other premium content.

Upload your 3-5 best sources into NotebookLM. Not 20. Three to five, chosen carefully. Quality and specificity beat volume every time.

Then immediately ignore the temptation to click "Summarize."

Summaries are passive. You read them. You feel informed. You forget them. Ask questions that force synthesis instead:

Each question forces the material to do something - surface a tension, prove a point, connect to something unexpected. That's processing. That's what summaries don't give you.

I tell my students: if your question could be answered by a single paragraph lifted directly from the source, it's too easy. Real comprehension lives in the spaces between sources, not inside any one of them.

Richard Feynman had one rule for understanding: if you can't explain something simply, you don't understand it yet.

Not because simple explanations are the goal. Because the act of explaining reveals exactly where your understanding breaks down. Every hedge, every vague phrase, every moment you reach for jargon when you can't find the plain version - those are the gaps.

I have run this test on myself dozens of times on topics I thought I understood. It is never comfortable. It is always useful.

Here's how to do it with Claude:

The first time you run this on something you've been reading, the feedback is uncomfortable. Gaps you didn't know existed appear immediately. Concepts you thought you understood turn out to be words you've been using correctly without knowing what they mean.

That discomfort is the learning. Gaps you can see are gaps you can fix.

Run the full sequence three times on the same topic. Address every weak point Claude identifies before the next pass. By the third explanation, you actually know the material.

I use Typeless for voice-to-text during this step. Explaining out loud is faster than typing and more honest - it's harder to hide behind hedged language when you're talking in real time.

Why this works: Research confirms that the harder your brain has to work to retrieve a memory, the stronger that memory becomes. The Feynman Test forces maximum retrieval effort on exactly the concepts you've processed least deeply.

You don't finish this system when you feel like you understand something. You finish when you've done something with it.

Three options in order of difficulty:

Listen on the move. Generate a NotebookLM Audio Overview of your sources. Customize the focus first:

I listen to these on morning walks with Alfie. Forty minutes of reinforcement that would otherwise be lost. By the time I'm back at my desk, the material has had another pass through my head without touching my work time.

Write the brief. Write one page summarizing what you've learned for a non-expert colleague. Then:

The writing organizes your knowledge. The Claude feedback shows you where the organization broke down.

Teach it to a real person. The highest-difficulty test and the most reliable one. If someone can follow your explanation and ask you good questions from it, you've learned it. If they're nodding politely and lost, you haven't.

I assign this to my students on every major topic. The ones who can teach it pass the exam. The connection is direct and consistent.

Day 1: Map prompt. Get the overview. Choose your 3-5 best sources.

Day 2: Load into NotebookLM. Ask synthesis questions. Find the tensions and connections.

Day 3: Feynman Test. Three passes. Address every gap.

Days 4-5: Audio overview on a walk. Write the brief. Get Claude's feedback.

Days 6-7: Teach it or build something with it.

Seven days. Any topic. Actually learned.

If this changes how you engage with the content you read every week, send it to one person still re-reading the same article hoping it will stick this time.

That's all I'm asking :)

See you next week.