This week we’re going to use AI for two very practical, very human problems: first, researching a major purchase without getting buried in 43 open tabs, and second, writing a professional decline email that is honest, kind, and doesn’t take you an hour to wordsmith. Both are places where people in their 40s, 50s, and 60s routinely bleed time and mental energy. AI can’t make the decisions for you, but it can clear away a lot of the clutter so you can think clearly.

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

Use Case 1 of 2

Use Case 1: Researching a major purchase decision

On a Tuesday evening around 9:30 p.m., after the house finally quiets down, Dana is staring at her laptop with that glazed “I regret starting this” expression. She’s 56, a high school counselor, and her 12-year-old dishwasher just died. The repair person gently suggested it was time to replace, not fix.

Dana has spent three nights in a row reading dishwasher reviews. She’s narrowed it “down” to eight models from four brands, each with mixed ratings and contradictory comments. One reviewer says a model is “whisper-quiet,” another says it “sounds like a freight train.” Some people insist stainless steel interiors are a must; others say it’s overkill. Energy efficiency ratings, extended warranties, installation fees - she has a note file on her phone, a printed Consumer Reports page with old coffee on it, and five retailer sites open in her browser.

The real problem: she is not choosing between eight dishwashers. She is choosing between the feeling of “I did my homework and made a solid choice” and the fear of “What if I missed something and regret this for ten years?” That’s why it’s exhausting.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

Copy and paste this into your AI assistant, then fill in the brackets with your real details and links.

> I want help making a major purchase decision. Please act as a calm, unbiased research assistant. > > STEP 1 - Understand my situation > Ask me up to 8 clarifying questions about my situation, priorities, and budget before you give any advice. Keep the questions short and plain-language. > > STEP 2 - Analyze the options I give you > After I answer your questions, I will paste in the products I’m considering with links and any notes I already have. > > For each option, create a short table with: > - Price (and note if it tends to be discounted) > - Key pros > - Key cons > - Any reliability or durability concerns mentioned consistently in reviews > - Anything that might annoy me day-to-day (noise, awkward controls, cleaning hassles, etc.) > > STEP 3 - Compare and summarize > Then: > - Compare the options specifically for someone in my situation and priorities (not a generic buyer). > - Tell me in plain language what tradeoffs I’m really choosing between. > - If you have a recommendation, give it, but explain it in terms of my priorities, not generic “best overall.” > - End with 3 - 5 questions I should ask the retailer or installer before I buy. > > First, ask your clarifying questions. Don’t analyze options yet. I will say “Here are the options” when I’m ready for you to compare specific models.

Why this prompt works:

This prompt forces the AI to slow down and ask questions instead of jumping straight to generic “best dishwasher 2026” advice. It tells the system exactly how to structure its analysis (tables, pros/cons, annoyances), which makes the information easier to compare. You’re not asking it to decide for you; you’re asking it to organize and translate. That plays to its strengths: pattern-spotting and summarizing.

AI will sometimes “fill in the gaps” if it doesn’t know something, especially about specific models or real-time prices. That can mean incorrect details. Always click through to the product pages yourself and verify specs, dimensions, and prices before you purchase. Think of AI as the friend who helps you sort your thoughts - not the friend whose word you take over the manufacturer’s website.

Use Case 2 of 2

Use Case 2: Writing a professional decline email

On Thursday afternoon at 4:15 p.m., Mark is sitting at his desk in his home office, staring at a blinking cursor. He’s 61, a financial planner who went part-time last year. An old colleague has emailed him:

“Hey Mark, we’d love to have you join our advisory board for a new fintech startup. Just 5 - 10 hours a month. Could be a great opportunity for you.”

Mark knows two things: he doesn’t have the bandwidth, and he doesn’t really want to tie his reputation to a high-risk startup. But he also likes this colleague and doesn’t want to sound ungrateful, dismissive, or old and tired.

So he does what many of us do: writes a sentence, deletes it, rephrases, worries it sounds cold, adds a softener, worries it sounds rambling, deletes it, checks email, opens the fridge, comes back, and is still stuck. A 90‑second task has become a 45‑minute exercise in emotional gymnastics.

This is where AI can be a surprisingly good collaborator - not to “fake” your voice, but to give you three or four clean starting drafts that you can edit into something that sounds like you.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

Again, copy/paste and fill in the brackets.

> I need help drafting a professional decline email. > > Context about me: > - My role/profession: [e.g., “I’m a part-time financial planner with a small solo practice.”] > - My general communication style: [e.g., “polite, concise, not overly chatty, but warm.”] > > Context about the request I’m declining: > - Who is asking and how I know them: [e.g., “an old colleague I like and respect, we worked together for 10 years.”] > - What they’re asking me to do: [brief description] > - My real reasons for saying no: [list 1 - 3 honest reasons, even if you don’t want to say all of them directly.] > - How firm my “no” needs to be: [e.g., “Very firm - no room for maybe later” or “Soft no - I might be open next year.”] > > Tone and constraints: > - Tone I want: [e.g., “warm, appreciative, but clearly declining”] > - Length: [e.g., “around 150 words or less”] > - I do NOT want: flattery that sounds fake, over-the-top apologies, or generic phrases like “due to unforeseen circumstances.” > > Please: > 1. Ask me up to 3 clarifying questions if anything is unclear. > 2. Then draft 3 different versions of the email: > - Version A: very brief and to the point but still polite. > - Version B: slightly warmer and more detailed. > - Version C: includes an offer to help in a smaller way (if appropriate, based on my reasons). > 3. Make each version sound like a real person in their 40s - 60s would write it, not corporate PR. > 4. At the end, list 2 - 3 sentences I could optionally add if I want to explain more, but keep them separate so I can ignore them if I want to stay brief.

Why this prompt works:

You’re feeding the AI the emotional and professional context it usually lacks: who this person is to you, how firm your “no” is, what you actually sound like. By asking for multiple versions with specific differences, you get options to react to rather than one “take it or leave it” draft. You also explicitly tell it what clichés to avoid, which is key to getting something that doesn’t read like a corporate form letter.

AI tends to over-apologize and over-compliment unless you rein it in, which can make you sound insincere or oddly deferential. Always read the draft out loud once - if a phrase makes you wince (“I’m truly humbled by this incredible opportunity”), delete or rewrite it. The AI doesn’t know your boundaries; you do. The goal is to sound more like yourself on a good day, not like someone else entirely.