This week we’re using AI for two very down‑to‑earth things: **spring cleaning with an AI‑made checklist** and **giving better feedback to a team member**. These are not “tech” topics; they’re everyday situations where a bit of structure makes life quieter, cleaner, and less stressful. AI happens to be very good at structure - if you ask it the right way.

Try at least one before Sunday. That is the whole assignment.

Use Case 1 of 2

Use Case 1: Spring cleaning using a checklist AI creates

On a rainy Saturday in early July, around 10:30 a.m., Linda stands in the doorway of her living room with a trash bag in one hand and a vague sense of dread in the other. She’s 62, works four days a week at a small accounting firm, and lives alone in a three‑bedroom ranch that’s quietly accumulated “stuff” over the last decade.

Her plan for “spring cleaning” this year is the same as it’s been for years: start somewhere, get overwhelmed, then give up halfway through and feel bad about it. She’s tried printed checklists from blogs, but they either assume she has a cleaning crew or want her to scrub the grout with a toothbrush. She wants something realistic: tailored to her house, her energy, and the fact that her knees are not what they were in 1998.

A friend mentions she used AI to build a one‑week room‑by‑room cleaning plan that felt surprisingly doable - declutter before deep cleaning, one area at a time, and specific tasks for each space. Linda decides to try it, but she has no interest in “prompt engineering.” She just wants: “Tell me what to do, step by step, and don’t kill me.”

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I want to do a realistic spring cleaning of my home, and I need you to create an organized, step‑by‑step checklist for me. > > Here are the details about my home and my limits: > - Home type: [describe your home: e.g., 3‑bedroom ranch with living room, kitchen, 2 bathrooms, small office, basement storage] > - Biggest problem areas: [list: e.g., cluttered surfaces, overstuffed closets, dusty baseboards, messy pantry] > - Time available: [e.g., 2 hours on weeknights, 4 hours on Saturday, none on Sunday] > - Physical limits: [e.g., not much bending or kneeling, no heavy lifting, okay with light dusting and wiping] > - Priorities: [e.g., want the main living areas and guest bathroom comfortable; closets and garage can be “good enough”] > > Based on this, create: > 1. A 7‑day plan, with each day focused on specific rooms or zones. > 2. A detailed checklist for each day with 8 - 15 specific tasks, ordered from “declutter” to “deep clean” (for example: start with removing items from surfaces, then dust, then wipe, then vacuum or mop). > 3. A separate list of “optional extra” tasks I can do if I have more energy (like washing curtains, cleaning behind the fridge, scrubbing grout). > 4. Short, encouraging notes for each day so the plan feels achievable, not overwhelming. > > Please format the output clearly with headings for each day (Day 1, Day 2, etc.), bullet points for tasks, and a checkbox symbol [ ] before each task so I can mark them off. Use plain language and assume I’m not trying to have a “perfect” house, just a comfortable, reasonably clean one.

Why this prompt works:

You’re giving the AI **specific constraints**: type of home, time, physical limits, and priorities. That stops it from producing a generic “ideal” checklist that doesn’t fit your life. You also tell it to **structure the plan** by day and by task order - declutter first, then clean - which is how many effective spring cleaning routines are organized. The request for “optional extras” and “encouraging notes” keeps the checklist flexible and psychologically manageable, instead of all‑or‑nothing.

AI doesn’t know your actual energy level on a bad day. It may still overestimate what you can do in an hour. If Day 1 looks like too much, simply reply: “That’s too ambitious. Rewrite Day 1 as no more than 8 short tasks I can finish in 90 minutes,” and have it adjust. Also, AI‑generated cleaning plans can include tasks that aren’t relevant to your home - like “clean ceiling fans” when you don’t have any - so read through once and delete what doesn’t apply.

Use Case 2 of 2

Use Case 2: Giving feedback to a team member

It’s Tuesday afternoon at 3:15 p.m., and Mark is staring at an email draft to one of his team members, Jasmine. He’s 48, runs a small consulting team of six, and has reached that point in his career where he’d rather do almost anything than have another uncomfortable feedback conversation.

The issue is simple but sensitive: Jasmine’s client reports have solid content but keep going out late and with small errors - typos, mislabeled charts, a missing attachment. It’s not a catastrophe, but it’s eroding trust with a key client. Mark doesn’t want to crush her; she’s smart, loyal, and clearly trying. But he also can’t keep quietly fixing things behind the scenes. He’s caught between “be direct” and “be kind,” and every sentence he writes swings too hard in one direction.

He’s heard that AI can help draft feedback emails, but he’s rightly wary of sounding robotic or formulaic. He wants his voice, his judgment, and his relationship with Jasmine to stay front and center. The AI’s job is to help him organize his thoughts and find phrasing that is both clear and respectful.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I need help drafting a feedback message for a team member. > > Here is the situation, in my own words: > - Role: I manage a small team of [describe your field: e.g., consulting, design, healthcare admin]. > - Person: [first name only], [brief description: e.g., smart and dedicated, 3 years on the team, generally well‑liked]. > - Issue: [describe the specific behavior, not the person: e.g., client reports are often delivered late and contain small errors, like typos and mislabeled charts. This has created minor frustration for a key client.] > - Impact: [explain why it matters: e.g., it makes us look less reliable, erodes client trust, and requires others to spend time fixing mistakes.] > - My intent: [e.g., I want to be clear about the problem, kind and respectful in tone, and focused on helping them improve rather than blaming.] > - Desired outcome: [e.g., agree on a specific plan to improve accuracy and timeliness over the next 60 days, with check‑ins.] > > Based on this, draft: > 1. A short email or talking points I can use in a conversation (about 3 - 6 paragraphs). > 2. Use a tone that is direct, calm, and supportive - no corporate jargon, no flowery praise. > 3. Include: > - One or two concrete examples of the issue (based on what I described). > - A clear statement of the impact on clients or the team. > - An invitation for them to share their perspective. > - 2 - 3 specific next steps we can agree on (for example, a checklist before sending reports, or a short buffer before deadlines). > 4. Write it in my voice: everyday language, no buzzwords, and assume I’m talking to an adult colleague I respect. > > At the end, give me 3 alternative sentence options for the opening line and 3 options for the closing line, so I can choose what feels most natural.

Why this prompt works:

You’re feeding the AI **your context and your intention**, not just “Write feedback for an employee.” That makes the output more grounded in your actual situation. You focus on behavior and impact rather than personality, which is the backbone of good feedback in any modern management advice. By asking for examples, an invitation for their perspective, and concrete next steps, you’re nudging the AI toward a constructive structure instead of a vague “you should improve” message. The request for alternate opening/closing sentences gives you flexibility to keep it sounding like you.

AI can easily drift into clichés: “I truly appreciate all your hard work” and “moving forward, I know you will rise to the occasion.” If the draft feels generic, treat it as a starting outline and rewrite key sentences in your own words. Also, AI doesn’t know your company culture or this person’s history beyond what you provide, so if the draft feels too soft or too harsh, ask it to adjust: “Make this more concise and more direct,” or “Soften this slightly while keeping the message clear.”