This week we’re looking at two very human problems that AI can quietly make easier: **spring cleaning using a checklist AI creates**, and **giving feedback to a team member** without losing your nerve or your clarity. Both are emotionally loaded: clutter and mess carry guilt, and feedback conversations can carry years of history. AI won’t do the hard parts for you, but it can structure the work so it stops feeling overwhelming.

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 late May, Denise, 57, stood in the doorway of her guest room and felt that familiar mix of annoyance and shame. Her daughter was visiting in two weeks. The “guest room” had quietly become the “random storage universe”: extra bedding, tax folders, two suitcases, a yoga mat, and a printer she meant to donate last year.

She’d tried downloading generic spring-cleaning printables before, but they always assumed a big house, three free weekends, and a personality that enjoys labeling bins. That was not Denise. She works full-time, gets home at 6, and lives in a 2‑bedroom condo with one creaky vacuum and a tendency to underestimate how long tasks actually take.

This time, instead of trying to design the perfect plan in her head, she opened an AI chat and said, “You figure it out. I’ll just tell you what my actual life looks like.” Ten minutes later, she had a **room‑by‑room spring cleaning checklist**, broken into thirty‑minute chunks, with clear priorities and realistic instructions. For the first time, the project felt like something she could follow instead of something she was already failing at.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I want your help to create a realistic, customized spring cleaning checklist for my home. > > Ask me 5 - 7 clarifying questions first, then build the plan. Do NOT assume you know my home or my schedule until you ask. > > Here is my situation to start from: > - Type of home: [ex: 2-bedroom condo, small yard, no garage] > - Who lives here: [ex: me and my partner, no kids, one shedding dog] > - My energy level and time: [ex: after-work only, 30 - 45 minutes a day, weekends are flexible] > - Deadline: [ex: I want things reasonably reset in 3 weeks, not perfection] > - Biggest problem areas: [ex: kitchen counters, paper piles, closets, guest room clutter] > - What I will NOT do: [ex: no moving heavy furniture, no ladders, no 8-hour cleaning days] > > After you ask your clarifying questions and I answer, create: > 1) A 7 - 14 day spring cleaning plan broken into daily “mini-missions” of 20 - 45 minutes. > 2) A room-by-room checklist that includes both obvious tasks (vacuum, dust) and commonly missed ones like baseboards, vents, light switches, remote controls, and under‑bed areas. > 3) A very short “bare minimum” version for days when I’m tired. > 4) A decluttering rule set for me to follow (for example, “one-year rule” or “20/20 rule”) so I don’t overthink every item. > > Make the checklist extremely practical and specific: > - Group tasks by room and by priority (high impact first). > - Estimate time for each task in minutes. > - Note anything that needs special supplies so I can make a small shopping list. > > At the end, summarize everything in one simple table with columns: Day, Area, Task, Est. Time, Supplies Needed.

Why this prompt works:

You’re not asking for “a spring cleaning checklist,” which would give you a generic internet list. You’re **forcing the AI to interview you first**, so the plan fits your home, your time, and your limits. You also specify formats: daily “mini‑missions,” a room‑by‑room checklist, and a table including time estimates and supplies, which makes the output something you can actually follow instead of vague inspiration. The decluttering rule set (like the “one-year rule” or “20/20 rule”) gives you built‑in decision shortcuts when you get stuck.

AI will happily assign you six hours of cleaning in a day if you don’t push back. It is bad at knowing how long things take in a real life with sore knees, phone calls, and low energy. After it gives you the checklist, read one day’s plan and **reality‑check the time estimates** against your actual experience; then ask it to revise: “That’s too much for a weekday. Reduce each weekday to no more than 45 minutes of tasks and move the rest to weekends.” Also, AI can forget your physical limitations (like lifting or allergies), so remind it if something looks unsafe or unrealistic.

Use Case 2 of 2

Use Case 2: Giving feedback to a team member

On a Tuesday afternoon, around 3:30, Marc, 49, stared at an email draft for the fourth time. One of his team members, Alisha, had been missing small but important details in client reports for months: wrong dates, incomplete numbers, forgetting to attach the appendix. None of it was catastrophic, but together it was eroding the team’s reputation.

Marc had tried “being nice” in passing comments: “Hey, careful on those dates,” and “Let’s double-check attachments next time.” Nothing changed. His boss had started to notice. He knew he needed a direct conversation, but every draft he wrote sounded either too harsh (“Your sloppiness is creating problems”) or too vague (“You might want to be a bit more thorough”). He didn’t want to crush her confidence, but he also couldn’t keep fixing her work at 9 p.m.

Instead of sending the awkward email, he opened an AI chat and said: “Help me write what I wish I could say, then help me turn it into something professional and kind.” What he ended up with wasn’t a script to read word‑for‑word, but a **clear structure for the conversation**, specific examples to use, and a short follow‑up email that captured the key points without drama.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I need help preparing to give performance feedback to a team member. > > First, ask me 6 - 8 clarifying questions about: > - My role and their role > - The specific behavior that’s a problem > - Concrete examples with dates or projects > - What I’ve already tried > - The outcome I want from the conversation > - Any sensitivities (health issues, past layoffs, personal stress, etc.) > > After I answer, do the following: > 1) Draft a **private conversation outline** I can use in a 20 - 30 minute 1:1. Use this structure: > - Opening (set a calm, respectful tone) > - Observations (specific facts and examples, no labels) > - Impact (why this matters for clients, team, or workload) > - Questions (invite their perspective) > - Agreement (what we each will do differently) > - Follow-up (how and when we’ll check progress) > 2) Suggest **3 - 5 sentences I can use verbatim** if I get stuck, written in plain, respectful language. > 3) Draft a short follow-up email summarizing the key points and next steps, in my voice, that I can send afterward. > > Adapt your tone to be: > - Direct but not aggressive > - Specific about behavior and impact, not personality > - Appropriate for a [choose: conservative / casual] workplace > > Very important: > - Do NOT tell me to avoid the issue or “soften it so much it becomes vague.” > - Help me be kind **and** clear. If I am too vague, point it out and suggest a more specific version. > - If my examples are weak (too old, too minor), tell me I need better examples before having the conversation.

Why this prompt works:

Feedback is highly contextual, so asking the AI to **interview you first** forces it to build the advice around your specific situation instead of giving generic management platitudes. The prompt also gives it a **clear structure** (opening, observations, impact, questions, agreement, follow‑up), which nudges it to think like a good coach rather than a quote generator. By explicitly stating tone (“direct but not aggressive”) and giving permission to push back if you’re too vague, you get something that feels more honest and less like corporate wallpaper.

AI doesn’t know your office politics, your HR policies, or your legal environment. It might suggest language that is fine in one culture and tone‑deaf in another. Before you use any phrasing, **run it through your own filter**: “Would I actually say this out loud? How would this land with this particular person?” Also, don’t let the polished script become a crutch; you still need to listen in the conversation. Use the outline as a guide, not a performance you’re trying to deliver perfectly.