(Let’s be honest, AI is better then you.)

I am sure you’ve heard that AGI (artificial general intelligence) is coming. Soon…

But stop for a second and just think… What is still exactly you still can do better then latest LLMs like Opus 4.8 or GPT5.5. or Fable 5 (r.i.p.)? I did such exercise and found out that:

AI writes better code than me. (python, js, sql, c++, assembler, you name it)
AI speaks more languages than me. (Chinese, Spanish, English, you name it)
AI designs better images than me. (try the new GPT image model, then try to argue)
AI writes better songs than me — and sings them too. (Any language. Any flow. Any mood.)
AI knows more than me. Biology. Chemistry. Physics. Math. Geography. History. Philosophy. Psychology. Electricity. You name it. Almost everything (‘almost’? not sure this word is needed here…)

Oh and yeah, It writes better articles than me. But not this one. This one's mine. Honestly (it did proofreading though). So maybe we already have AGI. Quietly. No announcement, no fireworks. Maybe..but this article is not about AGI…

Here's the thing.

A lot of people who try AI...still struggle.
They ask something. They get a bad answer. They think: "this doesn't work."

It works!

The issue isn't the AI. The issue is the prompt.

Garbage in. Garbage out.

Same model. Same day. Same task. One person gets magic, another gets garbage.

But its not a magic. it’s PROMPTING. Some people say the prompt is 50% of the answer.
I say it's 100%. Quality of answer = quality of your question.

So, in this article, I'm sharing with you one simple formula prompt for everyday use for any task… And one for when you're ready to go pro — a single prompt people call "the last prompt you'll ever need to learn."

(Why your best Friend is better then AI?)

So, what is a PROMPT?

A prompt is just your input. The thing you type before the AI replies. Any message (input) you type or dictate into chat is a prompt.

But even more… you're already prompting every time you text and ask your best friend "what should I get for dinner"… Basically it’s already a prompt.

Every time you ask anyone anything.

But.. there is some difference (yet) between prompting your Friend and AI.

Let's imagine a situation.

You text your best friend: "Bro, I wanna quit my job right now. I'm done."

Your friend doesn't open LinkedIn for you. He doesn't ask if you want help writing a resignation letter. He knows you. He knows you've said this exact sentence three times this year, always on a Thursday, always after a brutal week, and you've never once actually quit. He knows what you really mean is: "I'm overwhelmed, I need to vent, and I need someone to tell me I'm not crazy for feeling this way."

So he doesn't help you quit. He says "tell me everything," lets you rant for twenty minutes, lets get a beer together, and tells you to sleep on it.

He filled the gap. With history. With context. With actually knowing you.

Now type the exact same sentence into AI: "I wanna quit my job right now. I'm done."

It doesn't know about your three Thursdays. So it takes you at your word.

You'll probably get a resignation letter template. Or a checklist: "5 things to consider before quitting your job." Helpful, in a vacuum. Completely tone-deaf in yours.

You didn't want an exit plan. You wanted someone to hear you.

With AI, there's no friend on the other side. No history (yet). No tone read between the lines (yet).

Source: chatGPT 5.5

The quality of your input directly controls the quality of your output. That's the whole relationship. AI takes you literally. Word for word. Every time. At least, YET.

Vague question = Vague answer.
Lazy question = Lazy answer.
Specific question = Specific, useful, sometimes brilliant answer.

Your friend fills the gap with context he already has. AI fills the gap with whatever you actually gave it — nothing more.

So if you want it to stop handing you resignation letters when you just needed to vent... you have to tell it in the prompt to AI you have to call your friend! But all the other tasks, you should be able to resolve with AI and smart prompting!

(So, how to become better with prompts…)

Honestly, there are dozens of different "formulas" out there. Different structures, different gurus, different acronyms claiming to be the one true method.

But strip away the branding, and almost all of them boil down to the same thing: you need to give the AI these 6 elements in your prompt:

You won't remember all six elements every single time. You're not going to pull out a checklist before asking AI to help you write an email.

I know it, because I've bought professional prompt libraries before. I've built and sold my own. I've collected hundreds of "perfect" prompts for every possible situation. And I use them sometimes… 2-3% of all my prompts…

And in everyday life? I still don't use any of them.

What I actually use is one simple formula below: RISEN.

Prompt #1: ‘R I S E N’ - all you need

RISEN formula is easy enough to remember without looking it up. Flexible enough to cover almost everything I throw at it.

Created by Kyle Balmer, a prompt engineer who built his name teaching people exactly this skill. Five letters. Five steps. Covers almost anything you'll ever ask an AI.

R — Role

Tell the AI who to be. Not "you" — a specific expert.

Not: "Help me lose weight."
Instead: "You are a certified nutrition coach who specializes in sustainable fat loss for people who already train hard."

Same question. Completely different brain answering it.

I — Instructions

Tell it exactly what you want. Not vaguely. Precisely.

Not: "Help me lose weight."
Instead: "Build me a weekly meal plan to help me go from 176 lbs to 159 lbs. No fish. No nuts — I'm allergic."

S — Steps

Walk it through the process, like you would a new employee on day one.

"First, calculate my daily calorie and protein target based on my activity level. Then build a 7-day meal plan. Then give me a grocery list."

This is the step most people skip — and it's the one that turns a random answer into a structured one.

E — End Goal

Tell it why this matters. What does success actually look like?

"This should help me lose the weight without losing energy — I train 5 days a week and walk 10k steps a day, so I need fuel, not starvation."

Now the AI isn't just generating a diet. It's optimizing for an outcome that actually fits your life.

N — Narrowing

Set the boundaries. Length, tone, format, what to avoid.

"Meals should take under 30 minutes to prep. No fish, no nuts. Keep it realistic — I'm not trying to live on chicken and rice forever."

So, This is what it looks like once you combine all five:

"You are a certified nutrition coach who specializes in sustainable fat loss for people who already train hard. Build me a weekly meal plan to help me go from 176 lbs to 159 lbs. No fish, no nuts — I'm allergic. First, calculate my daily calorie and protein target based on my activity level (I train 5x a week and walk 10k steps a day). Then build a 7-day meal plan. Then give me a grocery list. This should help me lose weight without losing energy for my workouts. Keep meals under 30 minutes to prep, and keep it realistic — not just chicken and rice every day."

One paragraph. Thirty seconds to write.

And the gap between that and "help me lose weight" isn't small. It's the gap between a generic Google search and an actual nutrition coach sitting across from you, building a plan around your body, your schedule, your allergies.

So, I hope you will remember this one… I honestly use the idea of assigning ‘ROLE’ providing ‘STEPS’, and ‘End-goal’ or ‘Example’… and I noticed how different is the level of answers I get.

Prompt #2: ‘Professor Synapse’ — The Last Prompt You'll Ever Need

RISEN is what you use for quick, everyday tasks. One question, one answer, move on.

But sometimes you don't want a quick answer. You want a partner.

That's Professor Synapse.

Subscribe to keep reading

This content is free, but you must be subscribed to ScaleYourWeb Newsletter to continue reading.

Already a subscriber?Sign in.Not now

Keep Reading