This week we’re going practical and unglamorous: using AI to research a major purchase, and using AI to write a professional decline email when you want to say “no” without burning bridges. These are not flashy “change your life” topics. They’re the small, repeated decisions that quietly drain your time and energy - and where AI can give you back both.

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 night in late May, around 9:45 p.m., Denise was hunched over her laptop at the kitchen table. She’s 57, works as a clinical social worker, and has been putting off replacing her aging car for almost a year.

Her 2012 Honda has started making a grinding noise when she brakes. Her mechanic has gently said, “You can keep nursing it along, but at some point you’re throwing good money after bad.” Denise knows he’s right, but the last time she bought a car was in 2011. Now there are hybrids, plug‑in hybrids, fully electric, endless trim levels, and an avalanche of conflicting advice.

She has twelve browser tabs open: Consumer Reports, three car review sites, a Reddit thread, her credit union’s loan calculator, and two dealerships. Every article seems to introduce three new terms or options. She’s tired, and the thought of walking into a dealership underprepared makes her stomach knot. She doesn’t want to be an expert; she just wants to make a smart, defensible decision and move on with her life.

This is where AI shines - not by “deciding for her,” but by taking that pile of scattered information and helping her turn it into a clear, side‑by‑side comparison in plain English.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

Copy and paste this, then customize the parts in [BRACKETS] to your situation.

I want help researching a major purchase. Act as an objective guide, not a salesperson.

My situation: - I live in [CITY/REGION, COUNTRY]. - I am considering buying a [TYPE OF ITEM - e.g., compact SUV, home office laser printer, gas grill]. - My budget is roughly [PRICE RANGE, e.g., $25,000 - $32,000]. - I care most about: [LIST 3 - 5 PRIORITIES, e.g., safety, low maintenance costs, fuel efficiency, reliability, easy to use]. - I absolutely do NOT want: [DEAL-BREAKERS, e.g., complicated tech, high repair costs, poor customer support, unreliable brands].

Step 1: Ask me 5 - 10 clarifying questions to better understand my needs, including how long I plan to keep this item, how often I’ll use it, and any physical constraints (like garage size, weight, noise, etc.). Do NOT recommend anything yet.

Step 2: After I answer your questions, suggest 3 - 5 specific options that fit my needs, using a simple comparison table with: - Make/model (or exact product name) - Approximate price in my region - Pros (in plain language) - Cons (in plain language) - Why it might be a good fit for me

Step 3: For each recommended option, summarize: - Long-term costs I should consider (maintenance, repairs, subscriptions, energy use, etc.) - Any “gotchas” or common complaints from real users

Step 4: Ask if I want help preparing specific questions to ask a salesperson or expert, so I can verify this information in the real world.

Very important: - Be conservative and skeptical. If you’re not sure about something, say so and tell me what to double-check with a human expert. - Keep your language simple and direct. Avoid marketing language or hype.

Why this prompt works:

You’re telling the AI exactly who to be (objective guide, not salesperson), giving it your constraints (budget, priorities, deal‑breakers), and forcing it to slow down (clarifying questions first, recommendations later). The step‑by‑step structure prevents it from jumping straight to generic suggestions, and the comparison table makes the options easy to scan. Asking for “gotchas” and complaints pushes it to surface downsides instead of only shiny features.

AI will not know real‑time prices or local inventory, and it can be overly confident about specifics (like exact maintenance costs or regional availability). Treat the recommendations as a shortlist, not gospel. Always cross‑check a few key details (price, known issues, warranties) on up‑to‑date sites or with a human expert before you actually spend money.

Use Case 2 of 2

Use Case 2: Writing a Professional Decline Email

On a Thursday afternoon around 3:30 p.m., Malik stared at a blank email draft, feeling that familiar knot in his stomach. He’s 49, a senior manager in a nonprofit, known for being dependable and approachable - which means people ask him for things constantly.

A former colleague had emailed that morning:

> “Hi Malik! I’m launching a new community initiative and would love your help. Could you join our advisory board? It’s just a couple of meetings a month and some occasional feedback on documents. Your name would really help open doors with funders.”

The problem: Malik is exhausted. He’s already on two boards, his father’s health is declining, and his own team at work is stretched thin. He knows the honest answer is “no.” But he likes this colleague. He respects the project. And he does not want to sound ungrateful, aloof, or self‑important.

So he does what many of us do: he postpones the response. Tells himself he’ll “think about it.” The longer he waits, the worse it feels - and the more awkward the eventual reply becomes.

This is where AI can help you move from vague anxiety to clear, kind wording. The decision is still yours. The boundaries are still yours. But the AI can propose language you might not come up with when you’re tired, guilty, or worried about hurting someone’s feelings.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

You can use this for declining a board invitation, a project, a networking request, or anything similar by just changing the details.

I need help writing a professional decline email.

Context: - Who is writing: [YOUR ROLE, e.g., senior manager at a nonprofit, freelance consultant, department chair]. - Who I’m responding to: [HOW YOU KNOW THEM, e.g., former colleague, current client, acquaintance from a professional group]. - What they asked for: [BRIEF DESCRIPTION, e.g., to join their advisory board, to take on a new project, to speak at an event, to meet regularly as a mentor]. - My real reasons for saying no: [BE HONEST BUT BRIEF, e.g., too many commitments, health/family priorities, potential conflict of interest, doesn’t fit my core work anymore].

My goals for this email: - Be clear that the answer is no (not “maybe later”). - Be warm and appreciative of the invitation. - Preserve the relationship and keep the door open for smaller or more appropriate ways to help in the future (if that’s true). - Keep it to 2 - 3 short paragraphs, in plain language.

Please: 1. Ask me 3 - 5 clarifying questions about tone (formal vs informal), how well I know the person, and whether I want to offer an alternative (like a brief call, a referral to someone else, or a one‑time favor instead of an ongoing commitment). 2. After I answer, draft 2 different versions of the email: - Version A: More warm and personal, still professional. - Version B: More concise and businesslike, still polite. 3. In both versions: - State the “no” clearly and early. - Avoid over‑explaining or oversharing personal details. - Use language that doesn’t invite negotiation (no “maybe later when things calm down”).

At the end, briefly explain which sentences are doing the important boundary‑setting work, so I can adjust them confidently.

Why this prompt works:

You’re not just saying “write an email.” You’re feeding the AI the emotional and professional context: who you are, who they are, what they asked, and why you’re saying no. You’re also telling it what you care about (clarity, warmth, brevity) and forcing it to ask clarifying questions before drafting. Having two versions to choose from makes it easier to spot phrases that feel natural versus stiff, and asking for an explanation of the “boundary‑setting” sentences teaches you what to keep even if you tweak wording.

AI can drift into language that sounds either too corporate (“thank you for reaching out, per our previous correspondence…”) or too effusive (“I’m absolutely thrilled and honored!!”). You’ll need to read the drafts out loud and edit them so they sound like you. Also, be careful not to let AI talk you into offering more than you truly want to give; if it suggests adding, “Please feel free to reach out anytime,” ask yourself if you really mean that. If not, delete it.