This week we’re doing two very human, very practical things with AI: **spring cleaning using a checklist AI creates**, and **giving feedback to a team member** in a way that’s clear but kind. One makes your physical space less overwhelming. The other makes your work relationships calmer and more productive. Neither requires you to be “good with tech.”

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 Saturday morning in April, around 9:30, Maria stands in her kitchen staring at a jumble of mail on the counter, a sticky spot on the floor she keeps stepping over, and a dining chair draped in “temporary” clothes that have been there since November. She’s 57, works full-time, and jokes that her house looks “fine if you don’t open any doors or turn on bright lights.”

She’s tried printed spring-cleaning checklists before, but they always assume a spotless starting point and a burst of energy she doesn’t have. Her reality is 30 - 40 minutes here and there between work, her aging mom, and a knee that complains if she’s up and down the stairs too often. She doesn’t need a perfect home. She needs something realistic: small chunks, clear tasks, and permission to skip things without feeling like a failure.

Instead of searching for “the” right list, she opens ChatGPT and asks it to build a plan for her home, her schedule, and her limits. The result: a 4‑week checklist, broken into 30 - 45 minute tasks, with “must-do” versus “nice-to-do” labels and a short “if you only have 90 minutes” summary for busy weeks. Suddenly spring cleaning feels like a series of doable steps, not a vague mountain.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

Copy and paste this into ChatGPT or Claude:

> I want your help creating a realistic spring-cleaning checklist for my home that I can do in short sessions. > > Ask me 10 - 15 focused questions first about my home (number of rooms, trouble spots, storage space, how much time I have per day, physical limitations, etc.) so you understand my situation. > > After you have my answers, create: > 1) A prioritized checklist for a 4-week cleaning plan, assuming I can only work 30 - 45 minutes per day, 5 days per week. > 2) Tasks grouped by area (entryway, kitchen, bedroom, bathroom, living room, paper clutter, sentimental items, garage/basement/attic if relevant). > 3) For each task, include a plain-English description, a suggested time estimate, and the minimum tools/supplies I need. > 4) Mark tasks as “must-do” versus “nice-to-do” so I can skip things if I get busy. > 5) At the end, give me a one-page summary: “If you only have 90 minutes this week, do these 5 tasks.” > > Write everything in simple, everyday language. > Assume I’m not naturally organized and I get overwhelmed easily. > Keep the steps small and specific, more like “sort shoes into keep/donate/trash” than “organize the hallway.”

Why this prompt works:

This prompt works because it **forces the AI to understand your actual home and schedule before giving advice**, instead of handing you a generic list. You’re specifying time limits, grouping by area, and asking for small, concrete actions (“sort shoes into keep/donate/trash”) instead of vague instructions (“declutter hallway”), which makes it easier to start. The “must-do” vs “nice-to-do” labels and 90‑minute summary push the AI to prioritize, which is where most checklists fail.

AI doesn’t know your energy level on a bad day or how long it really takes you to, say, sort paperwork without getting distracted. It may underestimate or overestimate time, or suggest tasks that feel too ambitious. Treat the plan as a **draft**, not a decree: adjust time estimates, move tasks to different weeks, and delete anything that makes you tense just reading it. And double‑check any supply suggestions so you’re not buying a cupboard full of specialty cleaners you’ll never use.

Use Case 2 of 2

Use Case 2: Giving feedback to a team member

It’s Tuesday at 3:15 p.m., and David is rereading an email he’s about to send to Lena, a colleague who reports to him on a project. He’s 49, manages a small team, and has noticed that Lena’s last two client reports were late and full of small errors. The clients haven’t complained yet, but David can see the pattern.

He’s torn. On one hand, he needs those reports to be reliable; on the other, he doesn’t want to crush Lena’s confidence. The first draft of his email sounds harsh and slightly annoyed. The second draft is so softened that it barely says there’s a problem. He knows the usual advice - “be specific, focus on behavior, not personality” - but in the moment, with real people and real stakes, that’s easier said than done.

Instead of firing off the half‑baked email, David pastes his rough draft into an AI and asks it to help him sharpen the message: clearer examples, a calm tone, and a constructive path forward. He also tells the AI what he knows about Lena - her strengths, her past wins - so the feedback doesn’t read like a verdict on her entire performance. The end result is an email that plainly describes the issue, offers support, and asks for a plan, without sounding like a scolding parent.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

Copy and paste this into ChatGPT or Claude, then fill in the [brackets] with your details:

> I need help writing constructive feedback for a team member. > > Here is the situation: > - My role: [briefly describe your role, e.g., “I manage a small marketing team.”] > - Their role: [describe their role and responsibilities] > - What’s going well: [list 2 - 3 specific strengths or recent wins] > - What’s not going well: [describe 1 - 2 specific behaviors or outcomes that need to change, with examples and dates if possible] > - Impact: [explain how these issues affect clients, timelines, or the team] > > First, ask me 5 - 7 clarifying questions to make sure you understand the context (their personality, how long we’ve worked together, any relevant pressures they’re under, etc.). > > Then, write: > 1) A short script I can use in a live conversation (3 - 5 paragraphs, simple language, calm tone). > 2) A follow-up email that: > - Summarizes the key points from the conversation, > - Clearly describes the specific behaviors that need to change, > - Invites them to propose a plan or next steps, > - Reinforces that I value them and want to help them succeed. > > Use clear, everyday language. Avoid buzzwords and performance-review jargon. > Keep the tone firm about expectations but respectful and supportive toward the person. > > After you write the script and email, list 3 - 4 alternative phrases I can use if I want the message to sound slightly more direct, and 3 - 4 phrases if I want it to sound slightly more gentle.

Why this prompt works:

This works because you’re **feeding the AI the raw material it needs: your role, their role, what’s working, what’s not, and the impact**, instead of just saying “write some feedback.” That makes the output specific to your situation, not a generic HR memo. Asking for clarifying questions helps the AI understand nuances - how long you’ve worked together, their typical behavior - before it writes anything. And separating a **live script** from a **follow‑up email** matches how real feedback usually happens: first the conversation, then something in writing to make expectations clear.

AI can make feedback sound smoother than you naturally speak, which can be a problem if your team suddenly hears you using phrases that feel “corporate” or not like you. Before you use the script, read it out loud and adjust wording so it sounds like your voice. Also, remember that AI doesn’t know the full emotional history between you and this person; if there’s past conflict, your colleague may react differently than the AI assumes. Use its draft as a starting point, then layer in your judgment about timing, setting, and what this person can hear right now.