AI note taking workflow for lectures and readings

The goal of AI note taking is not a prettier transcript. The goal is a note that helps you answer future questions faster.

Step 1: Capture before cleanup

Write rough notes during lecture or reading. Do not ask AI to summarize something you have not processed at all.

Step 2: Structure the note

Ask AI to separate definitions, claims, examples, procedures, formulas, and questions. This turns raw material into a study surface.

Step 3: Add retrieval questions

For each major section, ask for a question that tests use rather than recognition. Keep only questions you could imagine on an exam.

Step 4: Verify sources

Mark anything not clearly supported by the lecture, reading, or textbook. A fluent note is not automatically a correct note.

Step 5: Compress weekly

Move the strongest questions and traps into a weekly review. Delete filler.

Prompt to clean up a messy note

Turn these rough notes into a study note.

Separate definitions, claims, examples, formulas, procedures, open questions, and exam traps. Use only the material below. For each section, add two retrieval questions and one likely confusion. Put anything unsupported or unclear in a "verify later" section.

What not to outsource

Do not let AI replace the first pass of listening, reading, or problem solving. The workflow works because you capture the raw material yourself, then use AI to structure it and expose weak spots. If the model adds facts that were not in class material, keep them in a separate verification bucket instead of mixing them into the note.

The highest-value output is usually not the summary. It is the list of retrieval questions, uncertain claims, and traps that tell you what to study next.

Use the prompt pack for the AI prompts and the weekly review template for the review cadence.