Planning a road trip with AI and writing a complaint letter that actually gets results sound like two very different problems, but they share the same quiet frustration: you’re doing unnecessary mental heavy lifting that software can now reasonably help with. This week we’ll use AI not to “optimize your life” in some grand way, but to shave real time off things you already do - without turning you into 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 night at 8:15 pm, Mark and Lisa are sitting at their kitchen table in Columbus, Ohio. The kids are grown, the dog is asleep under the chair, and their September calendar has a rare four-day window with nothing on it. Mark has always wanted to drive down through Kentucky and Tennessee and end up in Asheville, North Carolina.

He opens his laptop, pulls up a map, and within ten minutes he’s stuck. If they leave Thursday morning, how far should they drive the first day? Where’s a town that isn’t just another interstate exit with the same three chains? Lisa wants one “really good local place” for dinner and at least one scenic stop that isn’t overrun with tour buses. Mark is toggling between six tabs: Google Maps, three travel blogs, hotel sites, and a weather forecast. After twenty minutes, the fun part of “Hey, let’s take a trip” has turned into logistical soup.

This is exactly the sort of problem AI is good at: taking your fuzzy preferences - “not too rushed, some scenery, decent food” - and turning them into a concrete plan with options for you to tweak. You still decide where to go. The system just does the heavy lifting and gives you a draft.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

I’m planning a 4-day road trip and I want you to act like a practical travel planner who understands driving fatigue, realistic timing, and non-touristy stops.

Details: - Starting point: Columbus, Ohio - Ending point: Asheville, North Carolina - Dates: Leaving Thursday morning, returning Sunday evening - Travelers: 2 adults in their 50s, reasonably active but not hikers - Preferences: - No more than 5 - 6 hours of actual driving per day - Avoid super-touristy spots and crowded bus-tour attractions - At least one scenic stop each day (viewpoint, small town, interesting landmark) - One locally-owned restaurant recommendation per day (not a chain) - Lodging in safe, walkable areas (mid-range hotels or B&Bs, nothing fancy)

Please: 1. Propose a day-by-day itinerary with: - Departure and arrival times - Estimated driving time and approximate miles - 1 - 2 specific scenic stops per day (brief description of why they’re interesting) - 1 - 2 specific towns you recommend for overnight stays, with reasons 2. For each overnight stop, give: - 2 - 3 hotel or B&B suggestions (describe the type, not just the name) - 2 - 3 local restaurant suggestions, with the kind of food and approximate price range 3. Include a short “if the weather is bad” backup option for each day. 4. At the end, suggest 3 questions I should ask you to refine this plan further once I read your suggestions.

If there are multiple reasonable routes, briefly compare them and pick one to recommend, explaining why.

Why this prompt works:

You’re giving the AI enough detail that it can make sensible choices: starting and ending points, dates, driving limits, and what “good” means to you (local, scenic, not exhausting). By asking for a day-by-day itinerary, backup options, and follow-up questions, you’re nudging it to be concrete rather than vague. The “act like a practical travel planner” instruction steers the answer away from fantasy itineraries and toward realistic timing and tradeoffs.

AI will confidently recommend places that sound real but may be outdated or occasionally just wrong - restaurants that closed, viewpoints that are actually just highway pullouts. Before you commit, cross-check any specific hotel or restaurant names in a map app or review site. Also, AI has no sense of your personal risk tolerance; if safety is a concern, do your own quick check on neighborhoods instead of taking its word for it.

Use Case 2 of 2

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

Thursday at 3:40 pm, Denise is sitting in her home office in Phoenix, staring at yet another email from her internet provider. Two months ago, the company quietly raised her bill by $35 without warning. The service has been spotty, she’s spent hours on hold, and every “customer care agent” seems able to say only, “That’s our policy.”

Denise is not someone who flies off the handle. She’s a project manager by background, used to clear documentation and reasonable timelines. But this kind of corporate stonewalling makes her feel both furious and strangely helpless. She’s drafted three different complaint emails, and each one sounds either too emotional (“I’m sick of being treated like this!”) or too limp (“I’d like to request a review of my account, if possible…”). She wants a letter that is firm, factual, and persuasive - something that makes a supervisor decide, “We should fix this,” not “This can wait.”

This is where AI can help you find that sweet spot: calm but assertive, detailed but not rambling. You feed it the facts; it gives you a draft that sounds like the reasonable adult you are, just with better structure and cleaner language.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

I need help drafting a complaint letter that is calm, firm, and likely to get my issue resolved. Act like a consumer rights advocate who knows how customer service departments operate.

Here are the facts (please organize these clearly in the letter): - Company: [Insert company name] - Type of service: [e.g., home internet] - How long I’ve been a customer: [e.g., 7 years] - Problem: - My monthly bill increased from [old amount] to [new amount] starting in [month]. - I did not receive clear prior notice of this increase. - Since the increase, service has been [describe: unreliable, frequent outages, etc.]. - Attempts to resolve: - Number of times I’ve called or contacted support: [insert number] - Dates of contact: [insert dates if you have them] - What I was told: [brief summary of “policy” or non-answers]

My goals: - Clear explanation of the problem in plain language - Specific request: refund or credit for [time period], return to previous rate, or fair adjustment - A deadline for response (e.g., 10 business days) - Polite but clear indication that if this is not resolved, I will consider escalating (e.g., regulatory complaint, public review), WITHOUT sounding threatening or hostile

Please: 1. Draft the full letter as an email I can send to a supervisor or escalation department. 2. Use a respectful, professional tone, but be explicit about the impact on me (time lost, stress, disruption to work or home life). 3. Make the key facts easy to skim: brief paragraphs, clear timeline. 4. At the end, include 2 - 3 subject line options for the email that sound serious but not aggressive.

Write the letter in my voice, as an adult in my 50s who is reasonable but tired of being brushed off.

Why this prompt works:

You’re feeding the AI the raw material - dates, amounts, attempts to fix the issue - and asking it to do what it’s good at: structure, tone, and clarity. By specifying goals (refund, rate adjustment, deadline) and the desired voice, you prevent it from drifting into either legalese or generic “we value your business” fluff. The focus on skimmable facts makes it easier for an overworked supervisor to see that you’re organized and serious.

AI can sometimes overstep and sound more legalistic or threatening than you intend, especially when it mentions escalation options. Read the draft aloud before sending and tone down anything that sounds like you’re promising a lawsuit when all you want is a fair adjustment. Also, make sure all amounts, dates, and account details are accurate; AI won’t know if you mistyped something, and an incorrect fact can give the company an excuse to ignore the rest.