This week, we’re talking about **spring cleaning with an AI-made checklist** and **using AI to give better feedback to a team member**. On the surface, those sound unrelated. Underneath, they’re the same question: how do you take something fuzzy and exhausting (“clean the house,” “have a difficult conversation”) and turn it into clear, doable steps with a bit of help from a machine that doesn’t get tired or flustered.

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

Use Case 1 of 2

Use Case 1: Spring cleaning using an AI-created checklist

On a rainy Saturday morning, Linda, 58, is standing in her kitchen looking around at…everything. The cluttered pantry, the dusty ceiling fan in the living room, the mysterious sticky spot under the dining table. She’s been telling herself, “I need to do a deep spring clean,” for three weeks. But every time she starts, she wipes one counter, gets distracted by a junk drawer, then gives up when she realizes she’s only made a small dent and feels more overwhelmed than before.

Linda isn’t lazy. She’s busy. She works four days a week, helps her daughter with childcare one day, and her energy is not what it was at 35. Her real problem is that “spring cleaning” lives in her head as one giant blob. She can’t see the edges. She doesn’t know where to start, what order to follow, or how much is “enough.” She wishes someone would just hand her a realistic checklist, broken into manageable chunks, tailored to her house and her time.

That “someone” can be an AI - if she asks in the right way. Generative AI tools are surprisingly good at turning vague tasks into concrete plans and checklists, especially if you describe your home and your schedule.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I want you to act as a practical home organizer and cleaning coach for a woman in her late 50s with limited energy. > > Here is my situation: > - Home: [describe your home: number of floors, number of bedrooms/bathrooms, any special areas like garage, basement, balcony] > - Biggest problem areas: [list a few: e.g., “cluttered kitchen counters, overstuffed linen closet, dusty living room, messy home office”] > - Time available: [for example: “I have 2 hours on Saturday and 1 hour on Sunday for the next 3 weeks.”] > - Physical limits: [for example: “I can’t lift heavy furniture and I need to avoid working more than 30 minutes without a short break.”] > > Based on this, create a **realistic spring cleaning checklist** broken into **short sessions**. > > Please: > 1. Organize the checklist **by day and room/area**, in the order that makes the most sense (including decluttering before deep cleaning). > 2. For each day, give me: > - An estimated time (for example “about 45 minutes”) > - 5 - 10 **specific, checkbox-style tasks** (start at high surfaces and move down to floors). > 3. Include at least one “quick win” task in each session so I feel progress quickly. > 4. Make sure you cover often-forgotten spots (like baseboards, light switches, under sofas, inside fridge, tops of cabinets, shower heads, window tracks). > 5. At the end, summarize the **overall plan in 5 - 7 bullet points** so I can see the big picture before I start. > > Use plain language. Assume I am not using any task apps - I will print this checklist and mark it up by hand.

Why this prompt works:

It gives the AI **context** (home layout, time, energy limits) and **constraints** (short sessions, estimated time, specific tasks, forgotten spots), so it can build a realistic plan instead of a generic “clean everything” list. It also tells the AI how to structure the output - by day, by room, with checkbox-style tasks - so what you get back is something you can literally print and follow without editing.

AI does not know your house. It will sometimes suggest tasks that don’t apply (cleaning a dishwasher you don’t have, organizing a mudroom when you live in an apartment). Treat the checklist as a **draft**, not gospel: cross out anything that doesn’t fit, and add your own items where needed. Also, AI can be over-ambitious. If a “45-minute” plan looks like it would take you two hours, cut it in half before you start. The tool’s job is to make your life easier, not to shame you into professional housekeeper standards.

Use Case 2 of 2

Use Case 2: Giving feedback to a team member

Thursday afternoon, 3:30 p.m. Mark, 52, is staring at a blank email to Jasmine, one of his team members. Jasmine is smart, kind, and well-liked. She’s also been missing deadlines and handing in work with avoidable mistakes. Last week, a client pointed out three errors in a report; Mark ended up staying late to fix them himself.

Mark knows he needs to address this. His company has a culture of “we’re a family,” which sounds nice until you’re the one who has to have the uncomfortable conversation. He doesn’t want to crush Jasmine’s confidence or make her feel attacked. At the same time, he can’t keep quietly redoing her work. He’s caught between “be honest” and “be kind,” and he doesn’t have a script.

AI can’t have the conversation for him. But it can help him **draft clear, specific, respectful feedback**, so he goes in with language that is both human and firm. This is one of the most practical uses of AI: helping you say difficult things in a way that matches your values.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

> I want you to act as a manager-coach helping me give clear, kind, and direct feedback to a team member. > > Context (please reflect this in your answer): > - My role: [for example, “I manage a small team in a professional services firm.”] > - Relationship: [for example, “We have a good relationship; I like her personally, and I want to support her growth.”] > - Person: [for example, “Mid-career analyst, generally positive attitude, sometimes anxious.”] > - Situation: [describe in 4 - 6 sentences, factual not emotional: missed deadlines, specific errors, impact on clients, impact on you and the team.] > - My goal: [for example, “I want her to understand the seriousness, feel supported, and leave with 2 - 3 concrete action steps to improve.”] > > Based on this, draft: > 1. A short **opening paragraph** I can use in an email or meeting that: > - Acknowledges what she does well > - States the problem clearly, with 1 - 2 specific examples > - Focuses on the impact, not blame > 2. A **list of 3 - 5 bullet points** describing the key behavior changes I need (for example: double-checking reports, communicating early if a deadline is at risk, asking for clarification). > 3. A warm but firm **closing paragraph** that: > - Reinforces that I want her to succeed > - Makes standards and expectations explicit > - Invites her to respond or suggest what support she needs > > Please: > - Use simple, professional language appropriate for a person in their 30s or 40s. > - Do not be vague or overly “corporate.” Avoid clichés like “circle back” or “move the needle.” > - Keep the tone: respectful, honest, specific, and calm. > > After you draft the email text, add a short note to me explaining **why this wording works** (for example, how it balances empathy and accountability), so I can adjust it in my own voice.

Why this prompt works:

You’re feeding the AI **rich context** about the relationship, the specific behavior, and your goals, which keeps it from generating generic “HR-speak.” You’re also telling it exactly what format you want (opening, bullets, closing) and what tone to avoid (vague corporate phrases), so the draft is usable with minor tweaks instead of a complete rewrite. Asking for a note on “why this wording works” helps you learn the structure of good feedback, not just copy it.

AI does not understand your workplace politics or unspoken histories. If you just say “she’s been difficult,” it might write something too blunt or too soft. The quality of its draft depends heavily on how **specific and factual** you are in describing the situation. Also, never send an AI-written message untouched. Read it out loud. Remove any phrases that don’t sound like you, and double-check that the examples are accurate. If the AI accidentally minimizes a serious issue or makes it sound harsher than you intend, fix that before you hit “send.”