Most of the useful AI in everyday life looks painfully boring on the surface: cleaning your house and talking to people at work. No futuristic robots, no magic dashboards. Just slightly easier Tuesdays. This week we’ll look at two places AI can help in very practical ways: **spring cleaning using a checklist AI creates**, and **giving feedback to a team member** without tying yourself in knots or sounding harsh. Both matter because they remove friction from things you already know you should do - but don’t, because they’re annoying, emotional, or 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 **Saturday at 8:15 a.m.**, Maria stands in her kitchen, staring at a half-empty bottle of glass cleaner and a pile of mail that has quietly turned into a small mountain. She’s 57, works full-time, and has spent the last few years helping her aging parents downsize. Her own house, a three-bedroom place she’s lived in for 18 years, feels like a museum of unfinished decisions: the junk drawer that won, the guest room that became “temporary” storage, the pantry where mysteries go to age.

She knows she needs to do a serious spring cleaning, but just thinking about it makes her tired. The problem isn’t the work itself; it’s **figuring out where to start** and what, exactly, “spring cleaning” should include. Every checklist she finds online is either too basic (“wipe counters”) or assumes she has a staff and a three-story mansion. She doesn’t want a Pinterest-perfect home. She wants to feel less embarrassed when friends drop by and less guilty when she opens that hall closet.

So instead of trying to design the perfect plan herself, she decides to let AI do the planning - on her terms, for her house, and her energy level.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

I live in a [describe your home: e.g., three-bedroom house with one guest room, a small office, and a basement we mostly use for storage]. I want to do a realistic “spring cleaning” that I can spread over 2 - 3 weekends.

Please create a detailed, room-by-room spring cleaning checklist for me, using this structure: - Start with an overview list of all the spaces I should include. - For each room or area, list specific tasks grouped as: - “Must-do this spring” (high-impact, essential tasks) - “Nice-to-do if I have time” (optional tasks) - Make sure tasks are clear and concrete (e.g., “Empty junk drawer and sort into keep/toss/donate piles” not just “declutter”). - At the end, create a suggested schedule that breaks the work into blocks of 2 - 3 hours, spread over 2 - 3 weekends.

Here are the details of my home and situation: - Type of home and approximate size: [e.g., 3-bedroom, 2-bath home, about 1,800 sq ft] - Number of people living here and any pets: [e.g., two adults, one dog] - Problem areas I’m most concerned about: [e.g., overstuffed closets, cluttered kitchen counters, messy garage] - My energy level and physical limitations: [e.g., can stand and lift but get tired easily; no ladder work] - What “success” would look like for me: [e.g., less visual clutter, easier to find things, not ashamed to open doors when guests visit]

Use all of this information to make the checklist realistic for my situation, not an idealized version. If you need to make assumptions, choose the less ambitious option.

Why this prompt works:

This prompt gives the AI **specific constraints**: the type of home, your energy level, and your priorities. You’re not asking for “a spring cleaning checklist”; you’re asking for a **room-by-room plan** with “must-do” and “nice-to-do” sections, plus a schedule. That structure forces the AI to be concrete and practical instead of vague. Adding details about your problem areas and what success looks like helps the AI focus on what will actually make a difference for you, not for some imaginary showroom house.

AI has no idea how big your “junk drawer problem” really is. It might underestimate how long tasks take or suggest more than you can realistically do in a weekend. Treat the checklist as a **draft**, not a law. Cross things out, push items to “next month,” and adjust the schedule if it feels too ambitious. Also, ignore any task that requires equipment or effort you don’t have - if it suggests “steam-clean all carpets” and you don’t own a steam cleaner, downgrade that to “vacuum thoroughly” or remove it entirely.

Use Case 2 of 2

Use Case 2: Giving feedback to a team member

On **Wednesday at 3:40 p.m.**, David sits in his car outside the office, avoiding going back in for his one-on-one meeting. He’s 49, manages a small team at a professional services firm, and he’s good at his job - except for one part: **giving critical feedback**.

His team member, Jamie, is bright and well-liked, but keeps turning in client reports with sloppy formatting, inconsistent numbers, and vague conclusions. David has mentioned it once or twice, in passing, but never directly. He’s worried about hurting Jamie’s confidence, worried about sounding harsh, and honestly, a little worried about the conversation going sideways.

The result: David spends extra hours “fixing” Jamie’s work himself before it goes to clients. He’s quietly frustrated, Jamie doesn’t realize there’s a problem, and the pattern continues. He knows this is unsustainable and unfair - to both of them. David’s real problem isn’t knowing *what* the issues are; it’s finding the **right words and structure** to give feedback that is clear, fair, and actionable, without feeling like an attack.

Instead of agonizing alone, he decides to use AI as a private drafting partner - to help him say what needs to be said, in a way he’d be comfortable delivering face-to-face.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

I’m a manager preparing to give performance feedback to a team member. I want help drafting what I will say so it is clear, fair, and kind, but still honest and specific.

Here is the situation: - My role: [e.g., manager of a 6-person consulting team] - Their role: [e.g., associate consultant responsible for client reports] - What they do well: [brief list, e.g., strong client relationships, good ideas, reliable with deadlines] - The main problems I’m seeing: [specific issues, e.g., frequent errors in numbers, inconsistent formatting, vague conclusions in reports] - Impact of these problems: [e.g., I have to spend extra hours rewriting reports; it affects our credibility with clients] - What I want to happen next: [e.g., improvement in report quality, agreement on a plan, and regular check-ins]

Please: 1. Draft a short opening for the conversation that sets a supportive tone and explains why I’m bringing this up. 2. Write 3 - 5 specific feedback points that focus on behaviors and impact, not personality, each with: - What I’m observing - Why it matters - What “better” would look like 3. Suggest 2 - 3 questions I can ask them during the conversation to make it a dialogue, not a lecture. 4. Use plain, direct language I could say out loud; avoid corporate buzzwords.

After you draft the feedback, give me a brief explanation of why this wording is likely to land well.

Why this prompt works:

This prompt gives the AI **context** (your role, their role, what’s going well) and **specific problems** so it can avoid generic clichés. You’re clearly asking for **spoken language**, not an email, and you’re insisting on behavior-focused feedback with impact and examples. The structure - opening, specific points, questions - matches how a good feedback conversation actually flows, so the AI is forced to produce something you can imagine saying in order, not just scattered phrases.

AI doesn’t know your team member’s personality or history beyond what you share. It might produce language that feels slightly too formal, or too gentle, or too strong for your relationship. Read the draft out loud and adjust words so they sound like **you**, not like a committee. Also, remember this is your perspective; AI can’t see Jamie’s side of the story. Use the questions it suggests to genuinely invite their viewpoint, not just to move through a script.