Course resource
Personal AI Starter Kit
Your Week 1 deliverable. Work through it once and you will have a stack you actually use, rather than four subscriptions you forgot about.
Fill this in as you go. It is meant to be a document you keep, not one you read once.
1. Where AI should help you
Before choosing a tool, name the work. List five tasks you did in the last two weeks that were slow, repetitive, or annoying.
| # | Task | How often | Minutes each | Minutes/month |
|---|---|---|---|---|
| 1 | ||||
| 2 | ||||
| 3 | ||||
| 4 | ||||
| 5 |
Now mark each one:
- W — mostly writing or reading
- T — mostly thinking, deciding, or planning
- R — mostly repetitive steps across apps
- C — mostly creative or visual
W and T tasks are where a chat assistant helps immediately. R tasks are automation candidates for Week 5. C tasks come back in Weeks 4 and 6.
My top three AI opportunities:
2. Your baseline
Pick the single task at the top of your list. Do it once, the way you do it now, and time it honestly.
- Task:
- Time taken today:
- What made it slow:
- Quality, 1–5, judged by you:
Keep this. In Week 4 you will redo the same task with AI and compare. Without a baseline you will not know whether AI helped or just felt impressive.
3. Choosing your stack
Do not subscribe to anything yet. Use the free tiers for two weeks first.
The one-tool rule: pick one assistant and use it for everything for two weeks, even when another would be better. You learn a tool's real shape by hitting its limits, not by sampling four of them.
Score the tools you are considering against what you actually need:
| What matters to me | Weight (1–5) | Tool A | Tool B | Tool C |
|---|---|---|---|---|
| Quality of writing | ||||
| Handles long documents | ||||
| Searches the live web | ||||
| Works with my existing files | ||||
| Privacy and data controls | ||||
| Price | ||||
| Available in my country |
Current model names, context sizes and prices change constantly — check the live comparison rather than trusting a number written in a course. What does not change is the list of questions above.
My primary assistant for the next two weeks:
Why:
4. Setting it up properly
- Turn off training on your conversations (Data Controls / Privacy).
- Turn on two-factor authentication.
- Write custom instructions. Start with this and edit:
About me: [ROLE, INDUSTRY, WHO YOU WORK WITH]
I am usually trying to: [TYPICAL GOALS]
My audience is usually: [WHO READS YOUR OUTPUT]
How to respond:
- Get to the point. No preamble, no summary of my question.
- If my question is ambiguous, ask one clarifying question before answering.
- Say when you are unsure rather than guessing confidently.
- Use plain English. No corporate filler.
- For anything factual, tell me what you would check.
- Run the AI Safety Checklist.
- Put the app somewhere you will reach it without thinking — phone home screen, browser pin, keyboard shortcut.
5. Your first week
One task per day. Ten minutes each.
| Day | Do this |
|---|---|
| 1 | Ask it to explain something in your own field. Judge the answer — you will spot the errors an outsider would miss. |
| 2 | Give it your top task from section 1. Compare with your baseline. |
| 3 | Paste something you wrote and ask what is unclear. Do not let it rewrite. |
| 4 | Ask it to pressure-test a decision you are actually making. |
| 5 | Deliberately try to make it get something wrong. Learn its failure mode. |
| 6 | Redo Day 2's task, but write a better prompt. Note what changed. |
| 7 | Review: which of the seven would you do again? |
6. What to watch for
It sounds right and is wrong. The confident tone is identical either way. Verify anything you would be embarrassed to get wrong.
It agrees with you too readily. If you push back, it often folds regardless of whether you were right. Ask "is that actually true, or are you agreeing with me?"
You stop thinking. The goal is a faster version of your judgement, not a replacement for it. If you cannot explain the output, do not send it.
Long conversations drift. When answers get vague or it forgets earlier context, start fresh and paste in only what matters.
7. End of week review
- Minutes saved, honestly:
- What it was genuinely good at:
- What it was worse at than I expected:
- One thing I will do differently next week:
- Am I keeping this tool?
Bring this to the Week 2 session.