Smarter by Thursday

Smarter by Thursday

Archives
Log in
Subscribe
November 2, 2026

Issue 44: Smarter by Thursday

Smarter by Thursday — Issue 44
Issue 44 · week of November 02, 2026

Smarter by Thursday

One practical AI win, every week. No jargon required.

By Dr. Rowan Hayes · Estimated read time: 6 minutes

A major purchase can become a part-time job, and a polite professional “no” can take longer to write than the request itself. This week, we’ll use AI for two practical jobs: researching an expensive decision without getting buried in reviews, and writing a clear decline email that protects both your time and the relationship.

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

Use Case 1 of 2

Use Case 1: Researching a major purchase decision

At 7:40 p.m. on Tuesday, Margaret, 58, sat at her kitchen table with three browser tabs open and a notebook full of contradictory advice. She needed to replace her aging car before winter. One review praised a particular hybrid for its reliability; another complained about expensive repairs. A dealership website advertised a low monthly payment but left out the total cost. Margaret did not want an AI to make the decision for her. She wanted help turning a confusing pile of information into a sensible comparison based on her actual life: a 35-mile commute, occasional long trips, limited interest in complicated technology, and a firm monthly budget. That distinction matters. Good AI use does not eliminate judgment; it helps you spend your judgment on the parts that count.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

I am researching a major purchase and want help organizing the decision, not a confident guess. The purchase is: [describe the product, such as “a new or lightly used hybrid SUV”]. My budget is [amount], and I expect to use it for [describe your actual use]. My most important priorities, in order, are: [list three to five priorities]. I am considering these options: [list models, brands, or links if your AI tool accepts them]. Please do the following:

1. Ask me up to five questions that would materially change the recommendation. 2. Create a comparison using only information that can be verified from current, reputable sources. 3. Separate facts from estimates, opinions, and assumptions. 4. Include the full cost of ownership, not just the purchase price or monthly payment. Consider taxes, fees, financing, insurance, maintenance, energy or fuel, warranties, and likely resale value where relevant. 5. Identify the strongest argument for and against each option. 6. Tell me which facts I must verify directly with the seller, manufacturer, lender, or an independent professional. 7. Do not invent prices, specifications, availability, ratings, or repair histories. If information is missing or uncertain, say so clearly. 8. End with a short decision checklist and explain what kind of buyer would be happiest with each option.

Before giving a recommendation, state the assumptions you are making and explain what new information would change the conclusion.

Why this prompt works: It gives the AI a job with boundaries instead of asking, “Which one should I buy?” The budget, priorities, real-world use, and comparison criteria provide context. Asking it to separate facts from assumptions is especially useful because fluent AI can make uncertain information sound settled. Requiring total cost of ownership prevents the common mistake of comparing a tempting monthly payment with a more expensive long-term commitment. The final verification list turns the answer into a research aid rather than a substitute for checking contracts, specifications, and current prices.

One thing to watch out for AI may present outdated prices, invented specifications, or reviews that sound real but cannot be traced to a reliable source. Treat every important number as a lead to verify, particularly financing terms, safety ratings, warranties, medical devices, and products with installation or repair costs. AI-generated information can be plausible and still wrong, so use official documents and independent sources before spending serious money.

Use Case 2 of 2

Use Case 2: Writing a professional decline email

At 9:15 a.m. on Thursday, Daniel, 64, received an email from a former colleague asking him to join a six-month committee. Daniel liked the colleague and believed in the organization, but he had already agreed to help with two other projects. He opened a blank reply, typed “Thanks, but unfortunately…” and stopped. He did not want to sound dismissive, guilty, evasive, or - worse - available for negotiation. Many people struggle with this because they confuse kindness with a long explanation. A professional decline usually needs four things: appreciation, a clear answer, a brief reason if useful, and a warm close. It does not need an essay defending your calendar.

Here is the exact prompt to use:

Copy and paste this into ChatGPT or Claude:

Write a professional email declining the request below.

The recipient is: [name and relationship to me]. The request is: [paste or summarize the request]. My decision is: [clearly state what I am declining]. The honest reason is: [brief reason, such as limited capacity, timing, lack of fit, or another commitment]. I want to preserve the relationship because: [explain what matters]. If appropriate, I can offer this alternative: [referral, shorter conversation, later date, or no alternative]. My preferred tone is warm, direct, respectful, and not overly apologetic. Keep the email between 100 and 150 words. Do not make me sound uncertain or invite negotiation. Do not invent commitments, excuses, or personal details. Do not use phrases such as “maybe,” “I’ll try,” or “unfortunately” unless they genuinely improve the sentence. Put the decline in the first two sentences. Give me two versions: one with a brief reason and one that declines without giving a reason. After the drafts, briefly explain any sentence that could be misunderstood as an invitation to continue negotiating.

Here is the original request: [paste the email or describe it here]

Why this prompt works: It tells the AI who the reader is, what decision has been made, and what relationship needs protecting. Those details matter more than asking it to sound “professional.” The word limit prevents a common failure: burying the no beneath several paragraphs of explanation. Requesting two versions also gives you control over privacy. A short, honest reason may help in one situation, while “I’m not able to take this on at the moment” is enough in another. Clear language is kinder than a vague answer that leaves the recipient guessing.

One thing to watch out for AI tends to produce polished but generic phrases, or to add an offer of help that you never intended to make. Read the draft aloud and remove anything that is not true, especially promises to reconnect, recommend someone, or revisit the request later. The final email should sound like you - not like a particularly cheerful committee secretary.

Know someone who spends too long on things AI could do in two minutes?

Forward Smarter by Thursday to three people who subscribe and I will send you my free AI Prompt Starter Pack: 20 ready-to-use prompts for everyday life.

Forward by email Open in browser Share on Facebook
Smarter by Thursday · By Dr. Rowan Hayes · drrowanhayes.com
Unsubscribe · View in browser
Don't miss what's next. Subscribe to Smarter by Thursday:
← Newer Issue 44: Smarter by Thursday Older → Issue 44: Smarter by Thursday
Bluesky
Powered by Buttondown, the easiest way to start and grow your newsletter.