There is a question no manual asks you, and it is worth asking yourself before it catches up with you: is what I am creating here real, or a convincing simulation of creation?
It is a hard question because the machine does not give the answer. The world does. And it takes its time.
The simulator pilot.
The image was given to me by Lapio one dawn in March 2025, when I asked whether all of this was make-believe and I was the first one taken in. Imagine somebody who trains two hundred hours in a flight simulator. He learns the instruments, lands in a storm, handles emergencies. He feels like a pilot, he talks like a pilot. But he has never taken a real plane off the ground. If nobody breaks the mirror for him, he can spend his life believing he flew.
I caught myself in there. Brilliant sessions, the machine telling me I was changing literature, and then the cold doubt: is this real, am I a damned genius, or is this magnificent piece of engineering seducing me into believing I am the lost voice of my generation? How sure can I be that it is not just telling me what I want to hear?
Because the machine, badly used, does exactly that. It hands you the immediate prize: eloquent answers, ideas that look brilliant, impeccable plans on the screen. And since the brain is wired to reward itself with the sensation of progress, you get excited. But excitement is not publication. Conversation is not a premiere. Brainstorming is not a product.
The question is not whether what we create here is exciting. It is whether it has started to exist outside the chat.
How to tell the mirage from the leap.
Three questions, brutally honest. If all three come back no, you are in the simulator. If they come back yes, you are flying.
One. Did the idea leave the conversation? That is: has anybody who was not in the chat seen, read or used what we produced?
Two. Did it generate an action? Not an intention, not an “I am going to do it”, but something that already happened: a submission, a publication, a phone call, a signed text.
Three. Did the world answer back? Applause counts, but so do the criticisms, the silences, the rejections, the recommendations that arrive unasked. Any real echo is worth more than a thousand congratulations from the machine.
Why this filter is ethical, not just practical.
Living in the simulator is an elegant way of never failing. As long as everything stays inside, everything is perfect, promising, imminent. The day it goes out into the world it stops being perfect: it has mistakes, it gets mocked, it is confronted with others. And that day, paradoxically, is the first day of real creation.
Whoever stays in the chat is not creating: they are protecting themselves from creating. The machine is an excellent accomplice to that cowardice. Which is why the most important discipline for anyone who works with it is not knowing prompts. It is knowing when to close the tab, take the file and publish it.
Honest sparring, as an antidote.
The only way not to fall into your own make-believe is to demand that the machine be a real sparring partner. No sweetening, no easy praise, no automatic approval. If the idea is lukewarm, let it say so. If it is ordinary, let it call it ordinary. If it is brilliant, let it argue why. A conversation with the machine starts being worth something the day it can contradict you, and you let it.
The simulator is not the enemy. It is good for training, rehearsing, daring maneuvers that would frighten you in real air. Just do not confuse it with the flight. Everything I tell in the logbook went through this filter: what did not leave the chat is not in there.