04_Planning_the_Work — why does a good plan still fail, and what fixes it?¶
Level: 101 · for anyone whose to-do list for today is still half full at midnight
Techniques need time, and time needs a plan. This chapter has one lesson with a program, on why a day planned from honest estimates overruns almost every time, and two pages that gather the practical advice around it: what a goal has to look like to survive a semester, how to build a schedule and a to-do list, and what an AI is for when you learn.
| # | Page | The question it answers |
|---|---|---|
| 1 | The buffer hour | Why does a day planned from honest estimates overrun almost every time, and what does a reserved hour buy? |
| 2 | Goals and schedules | What makes a goal one you keep, and what should a week of study look like? |
| 3 | Learning with AI | What gets worse and what gets better when an explanation is one click away? |
The through-line¶
An estimate is a prediction, and the first chapter's habit applies: check it against what happened and correct the estimate, not the feeling. Lesson 1 shows why the correction is always upward, since task times are skewed, and why a plan that holds needs an hour in reserve. Page 2 takes the goal-setting research, which is among the best replicated in psychology, and the practical scheduling advice from Zakrajsek's chapter 3, marked by how sure each piece is. Page 3 is about the newest tool: an AI can make every technique in this library cheaper, and it can also do the homework for you, which is the most complete way there is to learn nothing.
Po polsku, w skrócie¶
Techniki wymagają czasu, a czas wymaga planu. Lekcja 1 pokazuje, dlaczego dzień zaplanowany z uczciwych szacunków prawie zawsze się nie mieści: zadania rzadko kończą się przed czasem, a często po nim, więc błędy się nie znoszą; ratunkiem jest godzina rezerwy i mnożenie szacunków przez własny, zapisany współczynnik. Strona 2 zbiera rady o celach (konkretne i mierzalne, tuż poniżej maksimum) i o tygodniowym planie. Strona 3 mówi, do czego służy sztuczna inteligencja w nauce: może potanić każdą technikę z tej biblioteki, ale może też odrobić za nas zadanie, co jest najskuteczniejszym sposobem, żeby niczego się nie nauczyć.
A note on the code¶
Lesson 1's task times are a model with the right shape, skewed to the late side, and its probabilities are exact. Pages 2 and 3 have no program; each claim on them says how strong its evidence is.