Terminology and Concepts to look out for
Content Corpus · sweep · model · persist · study preference prompt · AI Generated Output · study-guide budget
Relevance to Me
Thursday's midterm is open guide. Your guide is what you paste into this week's three forms. Today: Mid Term Prep Day 2 — the Content Corpus covers Sep 15 – 18. Monday's chunks are saved in Monday's form and already count toward your 3,000-word study-guide budget.
Your midterm-prep data processing pipeline — same three stages as Monday, new Content Corpus:
- Sweep the data. Open the Google Doc, read every ¶, and strike every ¶ you already know cold. What remains is the still-shaky data you actually need to study.
- Model the data. Split the remaining ¶s into four roughly-equal chunks, one per MuggsOfStudy tab. Add a study preference prompt (flashcards · quiz me · simpler words · an example · table it · trip me up). Generate one AI Generated Output per chunk. If a chunk came back weak, re-prompt with a different shape.
- Persist to your study guide. Submit each Chunk N AI Generated Output into item N of today's form. That is what gets printed on Thursday's midterm guide. An empty item prints as a blank — watch your 3,000-word study-guide budget.
Board Question
A chunk came back vague. Change the prompt, cut the input, or both?
Same three stages as Monday. New Content Corpus: Sep 15 – 18. MuggsOfStudy starts empty today — that is normal. Monday's form is saved.
- Sweep the data. Open Doc 2. Read every ¶. Strike what you already know cold.
- Model the data. Four chunks, one per MuggsOfStudy tab. Add a study preference prompt. Generate.
- Persist to your study guide. Chunk 1 → item 1, and so on. Submit.
Output vague? Cut the chunk smaller, or ask for a different shape ("quiz me on layer order"), then Generate again.
Work time starts now and runs until the exit ticket pops up. Goal: four filled items.