This is how to use AI to decide whether to repair or replace something

This is how to use AI to decide whether to repair or replace something.

A $260 repair can look sensible beside a $700 replacement until you learn that the repair covers one part, the item has failed twice, and the new option includes installation. The opposite can also happen: a dramatic quote makes replacement feel automatic even though one bounded repair could restore years of useful life.

Give AI the numbers you actually have and make missing facts visible. Do not ask it to diagnose a machine or promise remaining life. The job is to compare scenarios and produce the questions that would change the decision.

The repair-or-replace prompt:

My six-year-old canister vacuum needs a $210 motor repair with a 90-day repair warranty. A comparable new model costs $520 and has a two-year warranty. The current vacuum also needs a $35 hose soon, but its filters and tools are in good shape. I can manage one week without it.

Compare repair and replacement across two-year cost, downtime, warranty, known wear, and uncertainty. Show the arithmetic, do not estimate lifespan as fact, and list the three questions that could reverse the current lean.

Swap in real quoted totals and ordinary constraints. The scenario-and-reversal table is the part that matters.

Level up the ask:

One more move before you send it. Ask this first:

Rewrite my prompt above into a version that will get me a better result. Tell me what you changed and why, then answer the improved version.

Compare futures, not just today’s price

Ask for at least two plausible futures: the repair holds, or another major issue appears. A replacement also has more than one future because installation, accessories, consumables, or weaker repairability can add cost. The table should expose those branches without assigning fake probabilities.

• Known cost: Keep written prices, included work, and immediate extras separate.
• Unknown durability: Turn remaining-life guesses into questions for a technician or manufacturer.
• Daily friction: Include downtime, setup, disposal, and the value of keeping familiar accessories.

Do not let the table diagnose the item

If safety or the cause of failure is uncertain, stop at preparation. A qualified person and the manufacturer’s guidance decide whether use or repair is appropriate. Update every price and warranty from the current written source.

The useful result is not a universal percentage rule. It is a clear picture of what you know, what each path costs, and the next fact worth buying or asking for.

• Share only: Item age, observed behavior, public replacement price, quoted totals, and ordinary constraints; exclude addresses, serial numbers, service accounts, and payment details.
• Verify: Diagnosis, repair scope, warranty, safety, replacement price, disposal, installation, and expected life with the technician, manufacturer, seller, and written terms.

This is how to use AI.