Your First Week with AI: 7 Practical Exercises to Build Real Skills

Reading about AI is useful. Using it is better. Most people who say “I tried AI and it wasn’t that helpful” either gave up after one or two frustrating prompts or never moved beyond the obvious uses.

The exercises below are designed to do two things: build real skill with AI tools, and show you where they’re genuinely useful versus where they’re not. Do one per day. Each takes 15-30 minutes.

You can do all of these with ChatGPT (free tier), Claude (free tier), or Gemini. Pick one and stick with it for the week — you’ll learn more from going deep on one tool than sampling many.

Day 1: The explanation exercise

The exercise: Pick something you’ve always sort of understood but never fully grasped. Ask the AI to explain it to you.

Then push back. Ask follow-up questions. Ask it to use an analogy. Ask it to explain a specific part you’re still confused about. Keep going until you actually understand it.

Try: “Explain compound interest to me like I’m 16. Then give me a concrete example with numbers.”

Then: “I get the basic idea, but I don’t understand why the growth curve gets steeper over time. Can you walk me through that specifically?”

What you’re learning: How to have a learning conversation with AI — not just accepting the first response, but iterating until you actually understand something. This skill transfers everywhere.

Reflection question: What made the explanation click? What would have made it better?

Day 2: The editing exercise

The exercise: Take something you’ve already written — an email, a report section, a social post, anything — and ask the AI to improve it for a specific goal.

The key: be specific about what you want. “Make this better” is a bad prompt. “Make this more concise (cut 30%), clearer, and more confident in tone” is a good prompt.

After you get the revision, look at what changed. Do you like the changes? Do some feel wrong for your voice? Edit the AI’s version to bring it closer to what you want.

Try: Paste a work email you’re about to send and add: “Revise this email to be more concise (cut at least 25%), more direct, and to end with a clear single ask.”

What you’re learning: AI is a drafting partner, not a replacement for your judgment. The back-and-forth editing process — AI draft, human refinement — is the workflow that actually works.

Reflection question: What did the AI change that you’d keep? What would you change back?

Day 3: The brainstorming exercise

The exercise: Use AI for a brainstorm on a real problem you’re currently facing. Something work-related is ideal — a decision you need to make, a project you’re planning, an approach you haven’t settled on.

Give it real context. The more it knows about your situation, the more useful its brainstorm will be.

Try: “I’m planning [project]. The main challenge is [obstacle]. Brainstorm 10 different approaches I could take, including at least 2 that are unconventional. For each, note the main risk and main benefit.”

Then pick the 2-3 most interesting ideas and dig in: “Tell me more about approach #4. What would the first 3 steps look like? What am I not thinking about?”

What you’re learning: AI brainstorms are genuinely useful when they’re grounded in real context. You’re the decision-maker; AI is the thinking partner that generates options you might not have considered.

Reflection question: Did it suggest anything you’d actually use? What was the best idea?

Day 4: The summarization exercise

The exercise: Find a long article, report, or document you’ve been meaning to read but haven’t. Paste it into the AI and ask for a structured summary.

Don’t just ask “summarize this.” Ask for what you actually need: “Give me the 5 key findings, any caveats or limitations mentioned, and the main takeaway for [your specific role/context].”

Then test the summary by asking follow-up questions about specific points.

Try: Find any substantial article (1,500+ words) from your industry or field. Paste it and prompt: “Summarize this for someone who works in [your role]. Give me: (1) the 4-5 most important points, (2) any data or evidence cited, (3) the one thing I should do or think differently based on this.”

What you’re learning: AI summarization is good but not perfect. The quality depends heavily on the prompt specificity and on whether the document has clear structure. You’ll develop an intuition for when to trust the summary and when to spot-check against the original.

Reflection question: How accurate was the summary? Did it miss anything important? Did it include anything that felt off?

Day 5: The research kickoff exercise

The exercise: Use AI to get oriented on a topic you know nothing about. Not to get all the facts (remember: AI hallucinates), but to get a map of the terrain.

The goal: understand what the topic is, what the main debates are, what terms you’d need to know to do real research on it, and what questions you’d need to answer.

Try: Pick something in your field that you’ve heard about but don’t know well. Ask: “I know almost nothing about [topic]. Give me: (1) a plain-English overview, (2) the main things experts debate or disagree on, (3) 5 key terms I’d need to know to do deeper research, (4) 3 specific questions I should find answers to from primary sources.”

Then go find one primary source — an article, paper, or report — and read it. See how well the AI’s orientation matches what you find.

What you’re learning: AI is excellent for getting oriented; unreliable for getting the facts. The right use is to build the map, then fill it in with verified sources.

Reflection question: How useful was the AI orientation? What did you have to verify?

Day 6: The prompting exercise

The exercise: Take one task and try three different prompting approaches, comparing results.

Pick something simple enough to evaluate quickly — maybe “write a subject line for a promotional email about [something].”

Version 1: Simple ask. “Write a subject line for a promotional email.”

Version 2: Add context and constraints. “Write a subject line for a promotional email selling a $49 productivity app to busy startup founders. It should create curiosity, be under 50 characters, and not use the word ‘boost.’”

Version 3: Add a model/example. “Write a subject line for a promotional email selling a $49 productivity app to busy startup founders. Here are examples of subject lines I’ve liked in the past: [paste 2-3]. Write 5 options in a similar style.”

Compare the outputs.

What you’re learning: Prompting is a skill. More context and constraints almost always produce better results. Examples are often the highest-value thing you can add to a prompt.

Reflection question: Which version produced the best output? What made it better?

Day 7: The real project exercise

The exercise: Apply AI to something real you need to do this week. Not a demo — actual work.

Look at your task list. Identify something that involves writing, research, analysis, or explanation. Pick the one where AI could meaningfully help, based on what you’ve learned this week.

Use what you’ve practiced: give context, be specific, push back on outputs, iterate, verify facts that matter.

Some real-work applications that work well:

Draft an email or document, then refine
Prepare for a meeting by brainstorming what will come up
Break down a complex project into specific steps
Understand an unfamiliar concept for a presentation
Write the first draft of something you’ve been avoiding

What you’re learning: The workflow, not the tool. Every skill you’ve built this week — iterating on prompts, providing context, verifying outputs, using AI for drafts not finals — applies to everything you’ll do with AI going forward.

Reflection question: What would have been harder or slower without AI? What would you do differently next time?

What comes next

After a week of practice, you’ll have a sense of where AI is genuinely useful in your work and where it’s not. You’ll also have an intuition for prompting that you’ll keep refining.

The next level: read the prompting guides here on Zro2One — there are specific techniques (chain-of-thought, few-shot, role assignment) that will significantly improve your results. And the workflows guide is worth reading if you want to build AI into recurring work rather than just one-off tasks.

The skill builds quickly. Seven days from now, you’ll be noticeably better than you are today.