Some weeks, AI should save you time. Other weeks, it should save your sanity. This week is both. We’re talking about using AI to plan a **real-world road trip** without turning it into a spreadsheet project, and writing a **complaint letter that actually gets results** instead of getting ignored. These are very normal problems for very normal adults, and AI can help without you becoming “a tech person.”

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

Use Case 1 of 2

Use Case 1: Planning a road trip using AI

On Tuesday evening at 7:45pm, Denise is sitting at her kitchen table with a legal pad, a half-empty mug of tea, and three tabs open on her laptop. She’s 58, lives in Columbus, and wants to drive to Asheville next month with her husband, Mark. They like good coffee, easy hikes, and staying in places that don’t feel like beige conference hotels.

She’s trying to figure out: drive straight through or stop overnight? Where to stay? Are there any “must-see” stops that aren’t tourist traps? Every website feels like it’s trying to sell her something, and every suggestion is either for people with toddlers or Instagram influencers with unlimited time and money. After an hour, her notes say “Day 1: drive?” and “hotel??” and she’s wondering if this trip will even be worth the effort.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

I want help planning a realistic road trip. Please ask me a few clarifying questions first, then create a detailed plan.

Here is what I know so far: - Starting point: Columbus, Ohio - Destination: Asheville, North Carolina - Trip dates: October 10 - 14 - Travelers: me (58) and my husband (60) - Driving preferences: no more than 5 - 6 hours of driving in a day, prefer easy routes over scenic but stressful roads - What we enjoy: good coffee, easy to moderate hikes, interesting neighborhoods, small local restaurants, bookstores, live music that starts before 9pm - What we want to avoid: big chain restaurants, very crowded tourist attractions, very expensive luxury hotels - Budget: mid-range (we’re fine paying for quality, but don’t want “luxury resort” pricing) - Accessibility: no major mobility issues, but we don’t want steep, rocky trails

First, ask me: 1) Whether we want to stop overnight on the way there or drive straight through. 2) What kind of accommodations we prefer (small inn, Airbnb, standard hotel). 3) Whether we want a fairly full schedule or lots of free time.

After I answer your questions, create: - A day-by-day plan with driving times, suggested stops, and approximate arrival times. - 2 - 3 options for overnight stays each night with pros and cons. - 3 - 5 specific recommendations for coffee shops, hikes, and dinner spots that match our preferences. - A short note for each recommendation explaining *why* you chose it for us. - Any seasonal considerations for mid-October (weather, crowds, closures).

Please present the plan in clear sections by day, and keep the tone practical and realistic - not like a travel brochure.

Why this prompt works:

You’re not asking for “ideas,” you’re asking for a **structured plan** with clear constraints: driving limits, preferences, budget, and what to avoid. The first section tells the AI who you are and how you travel; the second section tells it how to respond, step by step. Asking it to **start with questions** forces it to fill in missing details rather than guessing. The request for pros and cons plus “why you chose it” produces more thoughtful, less generic suggestions.

AI will sometimes invent very confident-sounding details - like saying a café is still open or a trail is “easy” when it hasn’t checked recent information. It does not have live access to every local update. Always click through and confirm opening hours, current reviews, and any trail conditions before you commit. And remember: weather and traffic are unpredictable; treat the plan as a helpful draft, not a fixed itinerary carved in stone.

Use Case 2 of 2

Use Case 2: Writing a complaint letter that actually gets results

On Friday morning at 9:30am, Victor is on hold with his internet provider for the third time this month. He’s 64, retired, and does some consulting work from his home office in Phoenix. Last month, his internet went down three times during client calls. He’s been charged for a service visit that never happened, and his bill quietly crept up by $35 without explanation.

He’s tried the usual: phone calls, online chat, speaking to “supervisors.” Lots of apologies, nothing fixed. He knows he needs to put something in writing - an email or letter that says, “Here is what happened, here is what I expect” - but every time he starts, it comes out either too soft (“I’m a bit disappointed…”) or too angry (“This is outrageous and I’m furious…”). He wants to sound firm and reasonable, not like someone ranting on social media, and he wants an actual outcome: a corrected bill, a clear plan to fix the service, maybe a small credit for the trouble.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

I need help drafting a complaint letter that is firm, factual, and likely to get results. Please ask me a few clarifying questions first, then write the letter.

Here is the situation in plain language: - Company: [insert company name] - Type of service: home internet - Problem: repeated outages during the past month, incorrect billing (charged for a service visit that never happened), unexplained price increase - Attempts to resolve: multiple phone calls and online chats; apologies but no lasting fix - Impact on me: I use this service for paid consulting work from home. I’ve had disrupted client calls, wasted time, and stress trying to resolve this. - My desired outcome: corrected bill, removal of charges for services not provided, a clear plan to prevent future outages, and a reasonable credit for the disruption.

First, ask me: 1) The exact dates of the outages and any reference or ticket numbers from prior calls. 2) The specific amounts in question (previous monthly bill vs. current bill, amount of incorrect charge). 3) Whether I want the letter to sound more formal (like a traditional business letter) or more conversational but professional.

After I answer, write: - A clear subject line. - A short opening paragraph stating the purpose of the letter and the account details. - A concise timeline of what has happened (dates, issues, attempts to resolve). - A straightforward description of how this has affected me, focusing on practical impact rather than emotion. - A specific list of what I am requesting (for example: billing correction, credit amount, and confirmation of steps to fix the service). - A firm but polite closing that sets a reasonable timeframe for their response (e.g., 7 - 10 business days).

Make the tone: - Respectful but assertive - Based on facts, not venting - Organized and easy to skim

After you draft the letter, please suggest: - One sentence I could add or remove if I want to make it slightly stronger. - One sentence I could add or remove if I want to soften the tone.

Why this prompt works:

You’re telling the AI **exactly what kind of letter you want** - firm, factual, and outcome-focused - and you’re feeding it the story in plain language instead of legal jargon. Asking it to start with questions forces you to gather the dates, amounts, and details that make a complaint credible. The structure you specify (subject line, timeline, impact, requests, closing) keeps the letter clear and easy for a customer service rep or manager to process, which increases the chances of action rather than a canned apology.

AI can sometimes slip into exaggerated legal language (“This constitutes formal notice…” or “I demand immediate restitution…”) that sounds dramatic but is not helpful unless you truly intend to escalate. Read the letter carefully and remove anything that doesn’t sound like you or overstates your situation. Also, remember that no letter guarantees the outcome you want; you may still need to follow up, reference the letter in future calls, or escalate to a regulator or ombudsman if the company doesn’t respond.