Prompts That Decide

Most people use AI as a search engine with better manners. They ask it a question, get an answer, say thanks, and move on. That is fine for looking things up. It is a waste for the thing AI is actually best at, which is helping you think. The owners pulling real value out of these tools are not using them to find answers. They are using them to pressure test their own decisions before those decisions cost real money.

That is what today is about. Six prompts that turn any AI from a fancy lookup tool into the sharp advisor you cannot afford to hire, the one who pokes holes in your plan before the market does it for you. These are not prompts for writing copy or drafting emails. These are prompts for the decisions that actually move your business, the pricing change, the new hire, the big bet, the offer you are about to launch.

A quick word on how to use these. Do not just paste the prompt and accept the first thing the AI says. The value is in the back and forth. Give it real context about your situation, push back when it is wrong, and make it defend its reasoning. A workspace like Galaxy.ai lets you run the same decision past several models and see where they agree and where they split, which is its own useful signal. Treat the AI like a smart advisor who is sometimes confidently wrong, because that is exactly what it is.

The first and most useful prompt does the one thing you cannot do for yourself, which is argue against your own idea honestly. When you have made a decision, you are emotionally invested, and that investment quietly blinds you to the holes. The AI has no ego in your plan, so it can attack it cleanly if you tell it to.

The prompt is this. Tell it your decision, give it the full context, and say, your only job is to argue against this. Find every weakness, every risk, every reason this fails. Do not soften it and do not balance it with the positives. Make the strongest possible case that I am wrong. Then read the result not to dismiss it but to find the two or three objections that actually sting, because those are the ones worth fixing before you commit.

The magic here is that a good devil's advocate does not talk you out of the decision. It shows you the weak points so you can shore them up first. You walk into the decision with your eyes open instead of discovering the holes after you have already jumped. That single shift, finding the holes before instead of after, is worth more than any productivity hack.

A close cousin, and a favorite of careful operators. Instead of asking what could go wrong, you assume it already did. You tell the AI to imagine it is six months from now and the decision was a disaster, then work backward to explain why.

The prompt. It is six months from now and this decision failed badly. Walk me through the most likely story of how it went wrong, step by step, from the first crack to the full collapse. The reason this beats a normal risk list is that a story is more honest than a list. A list of risks feels abstract and easy to wave off. A specific narrative of how the thing fell apart makes the failure feel real enough that you actually take steps to prevent it.

Run this before any decision you cannot easily reverse. The good ones come back with a failure story that makes you slap your forehead, because the path to disaster was obvious in hindsight and invisible in the moment. The pre mortem just moves the hindsight to before the decision, where it can still do you some good.

Owners make a surprising number of decisions on vibes, then discover the math never worked. This prompt forces the math out into the open before you commit. You hand the AI the relevant numbers and make it model the decision honestly.

The prompt. Here are my real numbers. Walk through the math on this decision. Show me what has to be true for it to work, what the break even point is, and where the math gets uncomfortable. Do not be optimistic. Show me the version where I am being realistic, not hopeful. Then sit with the answer, especially the part where it tells you the assumption your whole plan depends on. Most decisions hinge on one number you have not actually checked, and this prompt surfaces it.

The AI will not have perfect data, so do not treat its output as gospel. Treat it as a forcing function. It makes you state your assumptions out loud and see them on the page, and half the time just writing the assumptions down is enough to reveal the one that does not hold. The number you were afraid to look at is usually the number that matters most.

If you want my full decision prompt vault, all six of these written out with the exact wording and the follow up questions that get the most out of each, that is loaded into the AI Workflow Blueprint at forty seven dollars. Reply BLUEPRINT and the vault is yours.

The opposite of the devil's advocate, and just as useful. When you are about to dismiss an option, the AI can build the strongest possible case for it, which protects you from rejecting a good idea for a lazy reason. We are all quick to wave off the path we do not want to take. The steel man makes you face the best version of it first.

The prompt. I am leaning against this option. Before I reject it, make the strongest, smartest case for why it is actually the right move. Assume a very intelligent person believes in it and explain their reasoning. Then check your own resistance. If the steel man is weak, you were right to pass and now you know why. If the steel man is strong, you just caught yourself about to make a mistake out of bias rather than logic. Either way you decide on the merits instead of the mood.

Most decisions get judged on their first result. The sharp ones get judged on what happens next, and then what happens after that. This prompt drags the downstream effects into view, the consequences of the consequences that quietly determine whether a smart looking move was actually smart.

The prompt. Here is my decision. Walk me through the second and third order effects. If I do this, what happens, and then what happens because of that, and then what happens because of that. Show me the chain, not just the first link. This is how you catch the price cut that wins customers now and trains them to wait for discounts forever, or the shortcut that saves an hour today and creates a mess that costs ten next quarter. The first order effect is obvious. The second and third are where the real outcome hides, and where good operators separate from impulsive ones.

The last one is less a single prompt than a practice. The more context you give the AI about your business, your goals, your constraints, and your past decisions, the better its counsel gets. A generic advisor gives generic advice. An advisor who knows your real situation gives advice you can actually use, so it is worth taking a few minutes to brief it properly.

The practice. Before asking for input on a decision, give it the brief. Here is my business, here is what I am trying to achieve this year, here are my constraints, here is what I have already tried, and here is the decision in front of me. Then ask. The difference in answer quality is dramatic. A tool like Claude can hold a lot of that context and reason carefully through it, which makes it a genuine thinking partner rather than a vending machine for generic tips. The investment is the briefing. The return is advice tailored to your actual life instead of a stranger's.

You might be thinking that this is what advisors and mentors and mastermind groups are for, and you would be half right. A good human advisor is worth more than any prompt. The problem is access. The sharp ones are expensive, busy, and not available at eleven at night when the decision is keeping you awake. Most owners do not have a board of advisors on call. They have themselves, a spouse who is tired of hearing about the business, and a gut feeling.

The AI does not replace the great human advisor. It replaces the absence of one. For the long stretches when you have nobody good to think out loud with, having a tireless sounding board that will argue, model, and pressure test on demand is a genuine upgrade over deciding alone in your own head. The bar is not whether the AI is smarter than the best mentor you could find. The bar is whether it beats the conversation you would otherwise have with nobody, and it clears that bar easily.

There is also a speed advantage that matters more than it sounds. A human advisor you see once a month forces you to batch your thinking and wait. The AI lets you pressure test a decision the moment it shows up, while the context is fresh and before the pressure to just pick something pushes you into a lazy call. Fast, honest feedback in the moment of decision changes the quality of the decisions themselves, because you are reasoning when you are sharp instead of rationalizing after you are committed.

The owners who get this build a quiet habit. Every meaningful decision gets a few minutes with the advisor before it gets made. Not to outsource the choice, but to make sure the choice was stress tested by something other than their own hopeful brain. Over a year, that habit is the difference between a string of decisions you blundered into and a string you actually reasoned through. The compounding is invisible day to day and enormous over time.

Six prompts is plenty, so do not try to run all of them on every decision. Match the prompt to the moment. For a reversible call, the devil's advocate alone is enough. For the big, hard to undo decisions, run the pre mortem and the numbers check together, because those are the ones where being wrong is expensive. The skill is not memorizing the prompts. It is building the habit of consulting the advisor before you commit, instead of after you regret.

And keep the right posture throughout. The AI is not deciding for you. You are the operator. It is the sharpest, fastest, most tireless sounding board you have ever had access to, and it works at two in the morning when no real advisor would pick up. But it is sometimes confidently wrong, so you weigh its input the way you would weigh a smart friend's, with respect and with judgment. The decision stays yours. The thinking just gets a lot better.

Pick one real decision you are sitting on right now. Run it through the devil's advocate and the pre mortem before you sleep tonight. You will either commit with more confidence or dodge a mistake you were about to make, and both are wins you could not have gotten alone.

If you want me to be the advisor in the room for your next big decision, working through your real numbers and your real options with you, that is the AI Business Accelerator at ninety seven dollars. Reply ACCELERATOR and we will think it through together.

Jordan

The AI Newsroom | Jordan Hale | ainewsroomdaily.com