Spring cleaning and giving feedback don’t sound like AI topics, but they’re exactly where these tools shine in everyday life. This week, we’re going to let AI do two very human jobs: turning that vague “I should really clean the house” feeling into a concrete checklist, and helping you say the hard thing to a colleague or employee without blowing up the relationship.

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 Thursday night in late June, Maria stood in her hallway staring at a sagging shelf piled with tote bags, old mail, and a dog leash she thought she’d lost. She’s 62, works part-time, and lives in a three-bedroom house she’s been in for 20 years. The house isn’t dirty, exactly - just… tired. Closets that never quite got sorted. A guest room that’s really a storage unit.

She’d promised herself a “big spring clean” back in April, but every time she tried to start, she froze. Too many rooms, too many decisions. She’d make a little progress in the kitchen, then remember the garage, then feel overwhelmed and sit down with her iPad instead.

Her daughter mentioned using AI to create a cleaning plan, which sounded both silly and intriguing. Maria didn’t want a generic Pinterest list; she wanted something that actually fit her house, her energy, and her schedule. One Saturday morning, coffee in hand, she opened her AI app and tried what you’ll see below.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I want you to create a realistic spring cleaning checklist for my home that I can follow over the next 2 - 3 weeks. > > Ask me 8 - 12 clarifying questions first so you understand my house, my energy level, and my schedule. Then create: > - A prioritized list of rooms/areas in the order you recommend I tackle them > - A specific checklist of tasks for each room (bullets, not paragraphs), starting with decluttering before deep cleaning > - An estimate of how long each room will take for someone working at a moderate pace > - A suggestion for how to break the work into daily 30 - 60 minute blocks > > Important constraints: > - I am between 40 and 70 and I get tired if I overdo it. > - I have some mobility limits, so be gentle with tasks that require ladders or heavy lifting. > - I tend to lose momentum if the list is too long for each day. > > Format the answer with clear headings for each room, and use checkbox-style bullets (like “[ ] Dust surfaces”) so I can copy this into a notes app and check things off. > > At the end, give me 3 “quick win” tasks I can do today in 20 minutes or less to feel like I’ve started.

Why this prompt works:

You’re not asking for “a spring cleaning checklist” in the abstract. You’re telling the AI to interview you first, then customize the plan to your actual home, energy, and time. You’re specifying the *format* (checkboxes, headings, short time blocks) so what you get back is something you can paste straight into Notes or a task app and use as a real tool, not just read once and forget.

AI doesn’t know your actual house. It will forget that you have, say, an overloaded craft room or that your “office” is really a corner of the dining room unless you tell it. When it gives you the first version of the checklist, read it like you’d edit a recipe: add, remove, and rearrange tasks so it reflects your reality. Also, some tools underestimate how long things really take, so treat those time estimates as suggestions, not as proof you’re “slow” if you run over.

Use Case 2 of 2

Use Case 2: Giving feedback to a team member

On a Tuesday afternoon around 3:30, David sat at his desk with a knot in his stomach. He’s 55, leads a small consulting team, and has worked with his analyst, Priya, for three years. She’s smart, loyal, and thoughtful - but lately her client reports have been coming in late and half-baked, and he’s been quietly fixing them at night.

He knows he has to address it. He also knows Priya is going through some family stress, and he doesn’t want to sound cold or harsh. In the past, his “helpful suggestions” have sometimes landed as criticism, and he’s seen people shut down. So he does what many of us do: he postpones the conversation and tells himself he’ll handle it “after this project.”

Instead, he opens an AI assistant and decides to rehearse the conversation on-screen before he has it in real life. He wants help with the wording, but he also wants to stay human - clear, kind, and specific.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I need help drafting feedback for a team member, and I want it to be clear, kind, and specific. > > Here is the situation (please read this carefully): > - My role: [briefly describe your role and relationship, e.g., “I’m a project lead and she reports directly to me.”] > - Their role: [briefly describe their role, seniority, and usual strengths.] > - The problem: [describe 1 - 2 recent examples of the issue: missed deadlines, quality problems, communication issues, etc.] > - Impact: [explain how this is affecting clients, projects, or other team members.] > - Context: [note any relevant personal or business context, e.g., “She’s dealing with a family issue and I want to be supportive while still addressing this.”] > > Based on this, draft: > 1. A short opening script (2 - 3 sentences) I can say at the start of the conversation that sets a supportive tone. > 2. A clear description of the specific behavior that needs to change, using “when you do X, it causes Y” language instead of personal criticism. > 3. 2 - 3 questions I can ask to understand what’s getting in the way from her perspective. > 4. 2 - 3 concrete suggestions for how we might improve the situation together (e.g., check-ins, deadlines, resources), phrased as collaborative, not punitive. > > Write this in my voice: professional but conversational, like a manager who respects the person and wants them to succeed. Avoid buzzwords and corporate clichés. Keep it under 500 words so I can practice it. > > At the end, suggest 2 different closing lines I can use to end the conversation in a supportive way.

Why this prompt works:

You’re not asking the AI to “fix” the person or write a formal performance review. You’re feeding it the key facts and asking for a *script* organized around what good feedback looks like: concrete examples, impact, questions, and joint problem‑solving. You also constrain the tone (“professional but conversational”) and the length so what you get back is something you can imagine actually saying, not a stiff HR memo.

AI will sometimes default to very polished, generic language - lots of “moving forward” and “leveraging strengths.” If the draft sounds like something from a corporate training video, push back. Ask it to “rewrite this in simpler, more natural language, like two colleagues talking,” and then edit the draft yourself so it sounds like you. And remember: AI can’t see nonverbal cues, company politics, or your history with this person. Use the script as a starting point, not a script you must follow word for word.