S-Tier Prompt Writing

It'll ask you a thing or two on its own. That's the floor.

Let's be straight about something. Modern AI isn't clueless. Ask it for help and it'll often toss a question back before it answers. "What's this for?" "Who's it for?" Good. That's a start.

But look at what that actually gets you. A decent answer. A fine one. The kind of result that's just... okay.

Here's the only question that matters. Is okay why you're here?

You took this class. Not to be average at this. You took it because you want results that are genuinely good. And the gap between okay and good is never an accident. You have to go get it on purpose.

Average is what you get when you leave it to chance.

The average person types a request, takes whatever comes back, answers the one question the AI happened to ask, and calls it done. Average effort, average result. Fair enough.

But look at what they actually did. They let the AI decide how good the answer would be. They handed over the wheel and hoped.

Maybe it asked the right things. Maybe it asked one shallow question and moved on. Maybe it never asked the thing that actually mattered for their situation. They'll never know, because they never checked.

That's leaving your result up to chance. And you're better than that, or you wouldn't be reading this.

So you take the wheel. You run the questions.

Don't sit back and hope it asks the right things. Tell it to ask. And don't ask it for "enough to answer." Ask it for everything it needs to make this genuinely good.

Here's the line:

Feel the difference? You're not lowering yourself to "get a usable answer." You're setting a bar, out loud, and making it clear that bar with you.

It comes back with sharper questions than it would ever have asked on its own. Because now it knows the target isn't fine. The target is excellent.

One round of questions is the floor. Push past it.

Here's where most people quit too early. The AI asks a few things, you answer, it goes. Done.

You can do better. When it finishes asking, make it keep going.

It will. And those three are usually the ones that matter most. The little details that turn a good result into one that actually stands out. They're exactly the questions it never would have bothered with on its own, because on its own, it was aiming for fine.

You're not aiming for fine. So you make it dig.

It's not the same as going step by step.

Quick one, since the last lesson was close.

Step by step controls how the answer comes out. Slowly, one piece at a time. This happens before the answer even starts. It decides how good that answer is allowed to be in the first place.

One paces the answer. This one raises the ceiling on it.

Use it on anything you actually care about getting right.

You don't need this for trivia. "What's the capital of France?" is Paris. There's no bar to raise.

You pull this out for the things that actually matter to you. The essay that's getting graded. The plan you're going to follow for real. The message you only get to send once. Anything where the gap between okay and great is a gap you'll feel.

If it's worth doing, it's worth making the AI work for the good version.

See it in action

The default:

Your standard:

You answer. It builds a plan aimed straight at your weak spots, in the way you actually learn, on your real timeline. Not a summary for everyone. A plan for you.

Same AI. Same five minutes. A completely different level of result, because you set a completely different bar.