-
@ noahrevoy
2025-05-29 17:39:18People say AI can’t create anything truly original. They’re right, because creativity doesn’t come from machines. Machines have no senses, no instincts, no ability to act in the real world. They’ve never tasted danger, felt grief, or made a sacrifice. Creativity belongs to those who have lived life. It comes to men and women who know what they’re doing through experience.
AI is not a writer or an artist. It is a tool. A calculator that uses words instead of numbers to guess what you want to hear. It has no soul, no experience, no skin in the game.
If you want to create something real, you must be the mind and heart behind the words. You must put your own inner fire into it. The AI is just the forge that lets you work the letters into place.
What AI Can't Do, And What You Must
This is how I write with AI. This is my process.
Step One: Define the Mission
First, I decide:
-
What I’m writing (length, format, scope, tone)
-
What I want to say (thesis)
-
How I’ll prove it (supporting points)
Step Two: Attack Your Idea Before You Defend It
Then I open up a conversation with the AI. Not to write but to challenge my ideas.
I ask it to generate a falsification prompt, a structured attack on my idea. But this requires precision. AI is not designed to challenge you. It’s trained to agree with you, to reflect consensus. So you have to hack it.
You must write a prompt that forces it to become adversarial, not based on popular opinion, but grounded in first principles. And even then, you must define which first principles matter. Clarity here sets the boundaries of the entire piece. This is the most important prompt in the process.
Then I run that prompt in a second thread. I hunt for weak points. I keep running tests until my idea holds. Only then do I continue.
Step Three: Map the Consequences
Next, I ask for second- and third-order consequences. What happens if this idea is applied in real life? What breaks? What works?
But even here, I don’t take its answers at face value. I ask it to falsify its own guesstimations, because that’s what they are, a mixture of guess and estimation, with no grounding in lived experience.
I push it to argue against itself, to challenge the implications it just offered. And I only accept what survives both prediction and pressure.
Now the shape is clear.
Step Four: Begin the First Draft
Only then do I tell the AI to begin drafting.
I give it a long, detailed prompt. Part of it is recycled between posts: my grammar rules, my format, my tone, my values. Some of it I keep saved in a PDF and drop in. The rest is tailored to the post at hand, instructions I generate during the previous steps, shaped by my interactions with the AI. The goal isn’t perfection. The goal is a useful skeleton I can flesh out. It gives me something usable but bland that covers the main points.
Step Five: Speak the Edits
Then I speak.
I dictate edits. I tell it what to keep, what to rewrite and how, what it missed. I move pieces around until the best arguments sit in the best possible order. When I’m finished, the post is 90% my own words. The AI added glue, structure, flow, formating. Nothing more.
Step Six: Quality Control
Now I run my quality checks. I ask a stock set of questions:
-
Does the summary of my idea still make sense?
-
Is the grammar consistent and clean?
-
Am I repeating myself?
-
Are the tenses correct?
-
Does my argument still hold up?
This is especially important for long posts. The goal is polish. Polish means clarity, rhythm, logic, and precision. Nothing missed. Nothing wasted. No extra words.
I also lean on insights from Joshua Lisec’s (@JoshuaLisec) The Best Way to Write It course. It taught me how to spot my typical writing flaws, and now I hold myself to that higher standard. The longer the document, the more important this phase becomes. Quality control doesn’t just clean up the surface, it keeps the message on brand, sharp, structured, and true.
This Is Amplification, Not Automation
This is not automation. It is amplification.
Automation replaces intention. Amplification serves it.
Automation assumes the machine should think for you. Amplification sharpens what you already know.
The AI didn’t create the message. It didn’t understand the truth of it. It didn’t live the stories that forged it. What it did, at its best, was help me say it faster, clearer, sharper.
Amplification means I can reach more people with less friction. It means I can test ideas at scale, refine arguments quickly, and stay in the rhythm of writing without stopping to reformat every sentence.
It’s not slop. It’s not a shortcut. It’s strategic leverage. But only if you already have something worth saying.
Most people don’t. Even many prolific writers don’t. They recycle ideas from one audience and feed them to another, a younger, more naive crowd that’s never heard them before. They mistake that for creativity.
I’ll be open with you: I rarely invent new ideas. I don’t need to. I write about ancient truths, time-tested strategies, the hard-won lessons of life.
What I do is connect what’s always worked with what’s possible now. I explain. I teach. And in person, I read people. I see what’s missing in them, and I help them see it too. Then together we build the missing pieces.
That’s not something you can fake. It’s not something AI can simulate. You have to live in order to have something worth saying. So the AI gives most people nothing. But if you’ve lived, if you’ve suffered and built and learned, then it can give you speed, clarity, and consistency in communicating that to others. But only if you lead it.
Why Most People Fail with AI
Most people fail with AI because they expect it to do the thinking. It can’t. It can only organize what you tell it.
If your results are vague, weak, or generic, it’s because:
-
You aren’t clear on what you want
-
You don’t know how to communicate boundaries
The problem is not the machine. The problem is your executive function is lacking.
Nonfiction Writing Was Always a Team Effort
Many nonfiction writers have always worked this way: Surrounded by assistants. Speaking out loud. Sorting ideas in real time. Having the best material refined and tested by teammates.
This isn’t about fiction, which is often best created in solitude, slowly, by a single mind over many years. Fiction doesn’t have to be real. It has to feel real. But nonfiction must follow truth. It must be grounded in reality, tested, and safe to apply. People get hurt if it isn’t.
There’s also a difference between writing because it’s your job and writing because you have something important to share. I don’t write for a living. I write because I know things that can help people, and that knowledge demands to be passed on.
If writing is your job, of course you’ll spend much more time on it. You’ll hire editors, assistants, researchers. Or AI can help with that, it’s low cost leverage. But if writing is your service, your legacy, your offering to the world, then you must be efficient. You must learn how to dictate, delegate, and get the truth out. Faster and better.
That’s what I do. And AI is the best assistant I’ve ever had for the money.
Even Trump writes this way. He speaks. His team organizes, sharpens, packages, gives him feedback.
Now you can do the same. For only $20 a month.
But only if you lead. Only if you know who you are, what you want to say, and how to direct a team. Because managing an AI is little different than managing people. You must have a solidly established ego. You must know how to set boundaries and hold them. You must be an expert communicator. The same executive skills that make you a capable leader in life are the ones that let you get real value from AI. Without them, it will run you.
Because this is your team. And you’re the head of it.
Final Truth: Lead or Be Led
Different people are getting vastly different results from AI. Some are producing brilliant work. Others are spinning in circles. When that happens, the problem isn’t the tool. It’s the user.
Any time a group uses the same tool and outcomes vary wildly, the difference is in the people, not the technology.
AI is perhaps the most complex tool humanity has ever built. It’s going to change the world, just like the printing press, the internet, and smartphones did. Whether that change is good or catastrophic depends on the quality of the people using it.
People with weak egos, poor boundaries, and low agency will be swept away. They’ll lose track of themselves. They’ll be absorbed into someone else’s machine. Not because they’re stupid. But because they’ve never trained their minds to resist.
And they will blame the tool. They’ll say, "This time it’s different." They’ll panic, moralize, or make excuses. But it’s the same pattern we’ve seen with every new influence system: TV, radio, social media.
The real issue is agency. If you have it, you use the tool. If you don’t, the tool uses you.
You can’t stop AI. But you can rise above its dangers. You can train your internal architecture to remain sovereign, no matter how powerful the machine gets.
If you want to build that kind of fortitude, the kind that makes you immune to being used, I can help you. That’s what I do.
Lead your thoughts. Lead your message. Use your tools like a man who knows what he’s building.
-