Today’s two topics are very down-to-earth: **using AI to plan a road trip that doesn’t leave you exhausted and disappointed**, and **using AI to write a complaint letter that actually gets results instead of polite brush-offs**. These are not theoretical uses of AI; they touch travel, money, and your time - all things worth protecting.

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 a rainy Thursday evening, around 7:30 pm, Daniel (58) sat at his kitchen table with a legal pad, three guidebooks, and six browser tabs open. He and his partner were driving from Denver to Santa Fe and then looping back through some national parks. He’d been “planning the trip” for two weeks and somehow still didn’t know where they were actually staying on night three.

His pattern was familiar: search “best road trip Denver to Santa Fe,” copy a list, realize half the stops were out of the way, then fall into a rabbit hole of reviews. By the time he looked up, it was 11 pm and he’d made exactly one decision: “we should probably avoid that one motel with bedbugs.” He wasn’t looking for luxury - he just wanted a **realistic** route with reasonable driving days, one or two genuinely interesting stops, and no frantic last-minute hotel scramble.

What mattered to him wasn’t “optimizing his journey” in the tech-bro sense. It was not arriving at each destination tired, hungry, and slightly annoyed at himself. He wanted the road trip to feel like a break, not a logistics project.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

Paste this into your AI tool and edit the parts in [brackets] to match your situation.

> You are my **personal road trip planner**. I’m a [age]-year-old traveling with [who you’re traveling with - e.g., “my spouse,” “two teenagers,” “a friend with mobility issues”]. > > I want to plan a **driving trip** from **[starting city]** to **[ending city]** leaving on **[date]** and returning on **[date]**. We will be driving our own car. > > Here is how we like to travel: > - Maximum driving time per day: **no more than [X] hours of actual driving**. > - We prefer **[small towns / scenic routes / avoiding big cities / avoiding mountain passes / etc.]**. > - We enjoy **[nature / museums / historic sites / food / live music / quiet walks / easy hikes under 2 miles]**. > - Mobility or health considerations: **[none / can’t do stairs / need frequent bathroom stops / can’t walk long distances]**. > - Our nightly budget for lodging is **[amount in local currency]**, and we prefer **[motels / mid-range hotels / B&Bs / chain hotels for points]**. > > Your task: > 1. Propose a **day-by-day itinerary** from departure to return that respects the driving limit. > 2. For each day, include: > - Departure and arrival towns > - Estimated driving time > - 1 - 2 suggested stops or short activities that match our interests > - 2 - 3 lodging suggestions in the arrival town that fit our budget level > 3. Aim for **realistic timing** (no rushing from place to place). > 4. Avoid one-night stays every night. Where possible, suggest at least one **two-night stay** in a place that makes sense. > 5. Present the itinerary in a clear table with columns: **Day, From/To, Drive Time, Activities/Stops, Lodging Suggestions**. > > Before you give the final itinerary, briefly list any assumptions you are making (for example about average driving speed or typical check-in times).

Why this prompt works:

You are telling the AI **who you are, how you travel, and what constraints matter**: driving limits, budget, interests, and mobility. That’s exactly what human travel planners need. You also define the *format* (a table, specific columns), which forces the AI to be structured instead of giving you a vague paragraph of “top 10 things to see.” Being explicit about “no more than X hours” and “at least one two-night stay” keeps the itinerary humane rather than exhausting.

AI will confidently **make things up**: opening hours, hotel prices, even whether a scenic road is open in a given month. Use the AI’s plan as a **draft**, not a final answer. Cross-check drive times on a real map, confirm hotel options on actual booking sites, and verify that suggested attractions are open on the days you’ll be there. Also watch out for it sneaking in one very long driving day “to make everything fit” - if you see a 9-hour day, tell it to revise with stricter limits.

Use Case 2 of 2

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

On a Monday afternoon at 3 pm, Andrea (64) sat at her desk rereading the email from her airline for the fourth time. Her flight home from a family wedding had been cancelled after five hours of “just one more delay.” She paid for an airport hotel out of pocket, missed a day of work, and spent another $50 on airport food.

She politely filled out the airline’s complaint form. Weeks later, she received a cheerful non-response: “We apologize for any inconvenience and appreciate your understanding.” No voucher. No refund. No acknowledgment of actual costs.

Andrea isn’t someone who wants to “go viral” with a rant on social media. She wants to be treated like an adult whose time and money matter. She’d like the **right** kind of letter: firm, factual, and focused on a reasonable outcome. But she finds herself either too polite (“I understand these things happen…”) or too angry, which she knows won’t help.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

You can paste this and then fill in the bracketed parts with your real details.

> You are my **professional consumer advocate and letter writer**. I need help drafting a clear, firm, but respectful **complaint letter** that has a good chance of getting a practical resolution. > > Situation: > - Company: **[company name]** > - Product/service: **[what it was - flight, appliance, repair service, hotel, etc.]** > - Date(s): **[date of incident and purchase]** > - Order/booking/reference number(s): **[numbers if you have them]** > - What was promised: **[briefly describe what you reasonably expected, in plain language]** > - What actually happened: **[brief, factual description of the problem - include timeline, people you spoke to, and what you were told, without insults or emotional language]** > - Financial impact: **[extra costs you incurred - hotels, meals, replacement items, lost work, etc., with approximate amounts]** > - Non-financial impact: **[missed events, significant inconvenience, stress, etc., in one or two sentences]** > > What I have already tried: > - **[Any calls, emails, or chats you’ve had with the company so far, and their responses]** > > What I am asking for: > - Be specific and reasonable: **[refund of X, credit or voucher of Y, repair or replacement, fee waiver, etc.]** > > Please draft a **polite but firm letter** I can send to the company that: > 1. Uses a clear subject line that will make sense inside a customer service system. > 2. States the facts in order, without emotional language or exaggeration. > 3. Clearly connects the company’s policies or reasonable expectations to what went wrong, if possible. > 4. Clearly and specifically asks for the outcome I’m seeking. > 5. Uses a respectful tone but makes it clear I expect the issue to be addressed. > 6. Is no longer than one page. > > Format the output as a complete letter I can copy, with: > - Today’s date left blank for me to fill in > - My name and address placeholder > - Company’s name and address placeholder > - A subject line > - A closing that sounds natural for a person in their [age] to send.

Why this prompt works:

You give the AI **structured raw material**: what was promised, what happened, financial and non-financial impact, what you’ve already tried, and what you want. That prevents it from inventing facts and keeps the letter grounded and specific. Explicitly asking for “polite but firm,” “no longer than one page,” and “clear subject line” guides the tone and length, which is where many complaint letters go off the rails. You’re essentially telling the AI, “organize my story into something a corporate system will take seriously.”

AI can drift into **overly dramatic or legal-sounding language** (“I hereby demand…” “This is unacceptable and outrageous…”) that may make you feel good but won’t help the person reading it. If the draft sounds like a TV lawyer, ask it to rewrite in “calm, straightforward language suitable for a reasonable adult.” Also, AI doesn’t know the company’s actual policies unless you paste them in, so avoid asking it to cite specific clauses unless you’ve provided the exact text.