Summary  
Leveraging AI to generate first drafts, summaries, and automate repetitive transformations in code or documentation while retaining human oversight for all critical decisions and final edits.

General domain of usage  
Text processing

By 2026, most knowledge workers use AI tools daily — ChatGPT, Claude, Gemini, Copilot, something. Used well, they're a real productivity multiplier. Used badly, they cause specific problems we're only starting to understand.


## The Good Uses

**1. First drafts.** AI is excellent at the cold-start problem. Give it a rough outline, get a starting draft, edit from there. Editing is faster than drafting from blank.

**2. Summarizing.** Feed in a 30-page report, get a one-page summary, then read the original parts that actually matter.

**3. Explaining unfamiliar territory.** New domain, new acronyms — AI is a fast, patient tutor.

**4. Repetitive transformations.** "Rewrite this in three different tones." "Extract every action item from this transcript." "Format this list as a table." Work that used to take 30 minutes now takes 30 seconds.


## The Bad Uses

**1. Thinking.** If you ask AI "what should I do about X" and you take the answer, **you've outsourced the decision.** Decisions about your work, your life, your strategy belong to you. AI is a good tool for exploring options; it's a terrible tool for making them.

**2. Communication you actually care about.** AI-written emails to family. AI-written thank-yous. AI-written wedding speeches. People can tell, and it damages trust faster than typos ever did.

**3. Anything where you'd benefit from struggling.** Learning math, writing prose, debugging hard problems. **The struggle is the point.** AI removes the struggle and removes the learning along with it.


## Three Personal Rules

- **AI for outputs you'll edit, not decisions you should make**;
- **First draft from AI, every word that ships is yours**;
- **Once a week, do something hard without AI** — write something cold, debug without Copilot, think through a problem on paper. Keep the muscles.

The tool is a hammer. Excellent for nails. Don't use it on yourself.


Which of the following are recommended uses of AI tools according to the chapter?

A course for knowledge workers who are tired of productivity gurus, 5 a.m. routines, and Notion templates the size of a novel.

Most popular productivity advice makes people less productive and more burned out — because it is built around the cult of work, not around how the brain and body actually function.

This course is about sustainable productivity: doing what matters, not lying to yourself that “you can do 8 hours of deep work every day” — you can’t — and not burning out by the time you’re 40.


Deconstruction before construction. We unpack why hustle culture, 5-AM routines, motivation-chasing, and time-blocking-as-religion make most knowledge workers less productive and more burnt out — not more capable. Includes the most-missed concept in modern productivity: boreout, the burnout from being chronically under-used, which produces the same symptoms as overload burnout but gets none of the attention.


The practical layer. A simple 3-part system that handles capture, decisions, and execution. Plus the chapters most people skip: how to actually tame inbox/Slack/notifications, the truth about how many hours of deep work you can sustain (it's not 8), how to use AI tools without losing your own thinking, a 25-minute weekly review you'll actually do, and the single most under-practiced productivity skill — saying no.


What turns a 6-month productivity push into a 30-year career without collapse. Sleep, exercise, and food as the boring foundation that beats every system. Active recovery (not "self-care theatre"). A personal energy audit. Boundaries that don't damage relationships. How to survive real crunch periods without breaking. Closes with a 30-day reset plan and a yearly rhythm you can actually keep.


AI Tools Without Becoming Dependent

The Good Uses

The Bad Uses

Three Personal Rules