Your first useful task

This is the capstone of the course. No more theory -- you are going to do something real. We will walk through one practical task together, step by step: summarizing a long email thread and writing a reply. By the end, you will have completed your first genuinely useful AI-assisted task.

Why this task

We chose "summarize an email thread and draft a reply" for three reasons. First, almost everyone deals with email, regardless of their job. Second, it uses skills from every previous lesson: understanding what AI is (Lesson 1), knowing how it processes language (Lesson 2), writing a clear prompt (Lesson 3), and applying the trust filter (Lesson 4). Third, the result is immediately usable -- you can actually send the reply.

Let's do this.

Step 1: Pick your email thread

Open your email and find a thread that fits these criteria:

• It is at least 5-6 messages long (enough that summarizing it by hand would take a few minutes).
• It is not sensitive or confidential (you will be pasting it into an AI tool).
• You need to reply to it but have been putting it off.

If you do not have a long email thread, you can use a long article, a document from work, or a group chat conversation instead. The process is the same.

Got one? Good. Copy the full thread.

Step 2: Get the summary

Open your AI assistant (ChatGPT, Claude, Gemini -- whichever you prefer). Paste the email thread and write a prompt using the four-part recipe from Lesson 3:

"Here is an email thread between me and my team. Please summarize it in bullet points: what was discussed, what was decided, and what is still unresolved. Keep the summary under 150 words."

Then paste the thread below your prompt.

Notice the structure: Task (summarize in bullet points), Context (email thread between me and my team), Format (bullet points, under 150 words, organized by discussed/decided/unresolved). We skipped the example because the format instructions are specific enough.

Hit send and read the summary.

Step 3: Review with your trust filter

This is where Lesson 4 pays off. Read the summary and ask yourself:

• Did the AI capture the main points? Compare it to your memory of the thread.
• Did it miss anything important?
• Did it add anything that was not actually in the thread? (Watch for this -- it is a hallucination risk.)
• Are the "decided" items actually things the group agreed on, or is the AI being optimistic?

If something is wrong or missing, tell the AI: "You missed the part about the budget discussion in the third email. Also, we did not actually decide on the deadline -- that is still open." The AI will correct itself.

This review step takes 30 seconds and is the difference between using AI well and using AI recklessly. Make it a habit.

Step 4: Draft the reply

Now ask the AI to help you write the reply. Again, use the four-part recipe:

"Based on this email thread and summary, draft a reply from me to the group. The reply should: acknowledge the key decisions, ask for clarification on the unresolved items, and propose next steps. Tone: professional but warm, like a friendly colleague. Keep it under 200 words."

Read the draft. It will probably be 80 percent right and 20 percent not-quite-you. That is normal and expected.

Step 5: Iterate once

This is where most people stop -- and where you will go further. Take the draft and give one round of specific feedback:

"Good start. Make these changes: (1) Replace the opening -- I would never say 'I hope this email finds you well.' Start with something direct like 'Thanks for the thorough discussion, everyone.' (2) In the next steps section, add a specific deadline suggestion: end of next week. (3) Make the closing less formal."

The AI produces a revised draft. Read it. It should now sound much more like something you would actually write. Make any final tweaks yourself -- change a word here, adjust a phrase there -- and it is ready to send.

You just did it

Take a moment to appreciate what happened:

1. You had a long email thread you were putting off.
2. You got a clear summary in seconds instead of re-reading everything.
3. You reviewed it critically (trust filter).
4. You got a solid first draft of your reply.
5. You refined it to match your voice.
6. Total time: probably 3-5 minutes for something that would have taken 15-20.

That is AI used well. Not magic. Not autopilot. A partnership where the AI does the heavy lifting of drafting and structuring, and you bring the judgment, the voice, and the final check.

The five-step pattern

What you just did follows a pattern you can apply to almost any task:

1. Gather -- Collect the input (email thread, document, notes, data).
2. Prompt -- Give the AI a clear instruction using the four-part recipe.
3. Review -- Apply your trust filter. Check for accuracy and completeness.
4. Refine -- Give specific feedback to get the output closer to what you need.
5. Use -- Make final human edits and put the result to work.

This pattern works for summarizing documents, drafting messages, preparing presentations, organizing notes, and dozens of other everyday tasks. You will use it again and again.

Celebrate -- then build on it

You have now completed your first real AI-assisted task. That is a genuine milestone. Most people who "try AI" open a chatbot, ask a random question, get a mediocre answer, and never come back. You did something different: you used a structured approach, applied critical thinking, and produced a real result.

In the final lesson, we will take this further. You will build three reusable prompt templates for the tasks you repeat most often -- and design a daily workflow that makes AI a natural part of your routine. Head to Your daily AI workflow when you are ready.

---