This week we’re doing two very down‑to‑earth things with AI: **spring cleaning using a checklist AI creates**, and **giving feedback to a team member** without tying yourself in knots. One is about clearing your physical space so you can think more clearly. The other is about handling tricky conversations so you can lead without losing sleep.

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 Wednesday in late July, Karen, 57, is standing in her hallway staring at the mess that snuck up on her. Her grandson’s toys are mixed in with old Amazon boxes, the coat closet is a jumble of scarves from three winters ago, and the guest room has quietly become “the place where stuff goes to die.” She’s not lazy; she’s overwhelmed.

Every time she thinks, “I should do a big clean,” she hits the same wall: where do I start, how long will this take, and what am I forgetting? She’s tried printed checklists from blogs, but they assume she lives in a showroom, not a real house with real people and a bad knee. What she wants is someone to look at *her* home, *her* energy, and give her a sensible plan that won’t make her feel like a failure if she can’t scrub the baseboards by Friday.

That “someone” can be AI - if you tell it what you actually need.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I need a realistic spring-cleaning plan for my home. > > Please ask me 5 - 7 questions first to understand my situation. Then create a simple, not perfectionist, checklist and schedule. > > Here is what you should know before you ask your questions: > - My home has these areas: [list your rooms or spaces, e.g., entryway, living room, kitchen, bathrooms, bedrooms, office, garage, basement]. > - I have about [X] hours per day and [number] days total I can realistically spend cleaning in the next month. > - I want to start with the most *visible* and *high-impact* areas first, so the house feels better quickly. > - I have these physical limits: [e.g., bad knee, can’t lift heavy things, get tired after an hour]. > > After you ask your questions and I answer them, please create: > - A room-by-room checklist with 5 - 10 tasks per room, using plain language. > - A suggested order of tasks over days or weeks, starting with quick wins. > - A short decluttering step before deep cleaning in each room. > - A list of supplies I should gather first (keep it short and realistic). > - Several tasks that can be done in 15 - 20 minute blocks for days when I have low energy. > - A simple “keep, donate, throw away” decision rule for clutter. > - A final day for finishing touches and things people forget, like baseboards, vents, light switches, filters, and under‑furniture spots. > > Keep the language simple and encouraging. Do not make the checklist too long. If a task is optional, clearly label it as “optional.” > If anything I ask for is unrealistic based on my time and energy, please gently point that out and suggest a more doable version.

Why this prompt works:

You’re not asking for “a spring cleaning checklist” in the abstract. You’re telling the AI about your actual home, your time, and your limits, and you’re asking it to interview you first so it doesn’t guess. You specify format (room‑by‑room, simple language, short bursts of tasks) and built‑in guardrails (not perfectionist, label optional tasks), so the output feels humane instead of like a military operation.

AI will happily give you far more tasks than a normal human can do in a week. If the list looks exhausting just to read, push back. Type: “This is too much for me. Please cut this list in half and focus only on the highest‑impact tasks per room.” Also remember: AI doesn’t know your body. If you have pain or mobility limits, adjust any task that involves ladders, heavy lifting, or scrubbing on your knees, or ask it: “Rewrite this, avoiding tasks that require bending or kneeling.”

Use Case 2 of 2

Use Case 2: Giving feedback to a team member

Monday afternoon, 4:30 p.m. You’ve already had three meetings and answered more emails than you care to count. You’re thinking about the conversation you’ve been putting off with James, 42, on your team. James is smart and loyal, but lately his reports have been late and a little sloppy. A client hinted at it last week, and you felt that familiar sinking feeling: “If I don’t deal with this, it’s going to get worse.”

You’re old enough to have seen feedback conversations go sideways. You’ve worked with bosses who were either too harsh or endlessly vague (“We just need you to step up more”) and you don’t want to be either. You want to be direct without being cruel, and you want James to leave the conversation clear on what needs to change, not guessing. You also know he’s been dealing with a sick parent, and you don’t want to bulldoze through that reality.

AI can’t have the conversation for you. But it *can* help you draft what you’ll say, so you’re not winging it when the stakes feel high.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I need help planning a feedback conversation with a team member. > > Please ask me 6 - 8 clarifying questions first about the situation, my goals, and this person’s context. After I answer, create a short script I can adapt, plus 3 - 4 key bullet points I want to cover. > > Here is what you should know before you ask your questions: > - I am [your role, e.g., “a director managing a team of 6 analysts”]. > - The team member is [their role, seniority, and anything important about their personality]. > - The main issue is: [describe in 2 - 3 sentences, e.g., “reports are late and contain errors, which affects client trust”]. > - I want to be direct but respectful, and I don’t want to sound like a script or like corporate jargon. > - There are relevant personal factors: [e.g., caregiving, health, workload] that I want to acknowledge without making excuses. > > After I answer your questions, please provide: > - A 3 - 5 minute opening script for the conversation, written in natural, conversational language. > - Clear phrases I can use to describe the impact of the issue, not just the behavior (e.g., “when X happens, Y is the consequence”). > - 3 - 4 concrete examples I can mention so the feedback feels specific, not vague. > - 3 - 4 open-ended questions I can ask them to understand their perspective and what support they need. > - A short summary I can use at the end of the conversation to confirm next steps and expectations. > > Please avoid buzzwords and management clichés. Keep the tone calm, firm, and humane, like a seasoned manager who wants this person to succeed. > If anything I’m asking for sounds unfair or unclear, point that out and suggest a better way to phrase it.

Why this prompt works:

You’re treating the AI like a quiet, experienced drafting partner: first it interviews you to get context, then it helps you structure your thoughts. You’re asking for specific elements (opening, impact statements, examples, questions, closing), which forces the AI to turn vague “you should give clear feedback” advice into usable sentences you can practice. And you’re explicitly banning buzzwords, which keeps the language human and age‑appropriate instead of sounding like a TED Talk.

AI doesn’t know your company culture or the legal landscape. It might suggest language that sounds fine on paper but would land badly with your HR team or in your specific environment. If something feels off - too harsh, too soft, or too formal - trust your gut and adjust. Also, don’t let the script replace listening. The danger is you focus on “getting through your points” and forget to hear what the other person says. Use the script as a starting point, then let the real conversation breathe.