This is not just the story of how a text is born. It is the chronicle of a progressive alignment between a determined human mind and an artificial intelligence tool, forced to learn, to adjust its aim, to understand nuances that were not in its instruction manual. It is the story of how Gabriele Cripezzi shaped, phrase by phrase, the prologue of his book Those Who Believe in God End Up in Hell, using ChatGPT not as a content generator, but as a refining partner, to be educated and guided.
The Start: A Clash of Perspectives
It all begins with a simple message: "Help me write". ChatGPT responds with enthusiasm as a generic assistant, offering tones and formats. Gabriele's reply is an immediate downsizing: "I tell you my thought, you write it correctly". There is no room for creative interpretations or extraneous suggestions. The role is clear: faithful transcriber, not co-author.
Gabriele then presents his monumental thesis: a two-thousand-year plan, born from an agreement between Greek philosophers and Roman rulers, which used the Bible as a tool for human selection, leading to the current, indecent inequality and the next, inevitable Great Reset. The text is dense, repetitive in structure, deliberately obsessive in hammering home key concepts such as lies, philosophical slavery and hell on Earth.
ChatGPT's first response is a betrayal in good faith.
AI summarizes, synthesizes, "organizes" thought into a more fluid and narrative language. It falls into the trap that many listeners fall into: it believes it must improve the form, without understanding that, for Gabriele, form and substance are inseparable. The reaction is dry:
"No, you haven't explained it well. Maybe I need to use a more advanced model?"
This is the first, fundamental educational clash. Gabriele does not criticize the content, but the method. The AI has misunderstood its role.
The training phase: precision versus interpretation
ChatGPT understands the error and changes its approach. It no longer proposes rewrites, but asks targeted questions to define the scope:
"I'll ask you some targeted questions..."
It asks for clarifications on the ancient plan, the nature of the Bible, the concept of hell, the Great Reset. Gabriele responds with a dry "Both things" or "Yes", rejecting any binary simplification. The AI attempts a new rewrite, but Gabriele intervenes again, correcting the lexicon:
"Instead of humanity, talk about the peoples. The peoples lived under oppression. And don't use words like immersed, because they are terms that I don't like and don't suit my way of speaking."
The instruction is clear: no literary, evocative or "mystical" terms. The language must be concrete, direct, philosophically precise.
ChatGPT attempts a third approach, but Gabriele blocks it again, providing even more granular instructions:
"Try to start like this: the causes or the true causes of indecent inequality... Then try to make longer sentences... instead of talking about forces, say directly who you're talking about, without adding that they are forces. It sounds too mystical to me."
Every correction is a lesson in style and coherence. The AI is learning that every word is a brick in a precise ideological structure.
The Turning Point: Starting from Scratch
After three failed attempts, Gabriele makes a radical decision:
"No, I don't like it. I want to try telling you everything from the beginning again and then see what comes out. Maybe the problem is my starting message."
This is the key move. He rephrases the entire concept with a slightly different angle, more dialectical and less narrative. He talks about exploitation, rebellion, a collaboration proposed by philosophers as an alternative to war. The logical structure is the same, but the emphasis is more on the mechanics of the agreement than its epic nature.
This time, ChatGPT grasps the difference. Its rewrite is more adherent, but Gabriele is not satisfied. He provides surgical corrections:
"The Bible is not a meeting point. The Bible is the sanction of the agreement... Instead of talking about people capable of reading between the lines, let's talk about people who had the right key to reading..."
When you say a soul condemned to suffer, emphasize and specify condemned to suffer until the last of the days.
If guided with precision and firmness, it can become an invaluable refining partner, capable of resisting the temptation to embellish and adhering to a rigorous mandate.
The prologue of Those Who Believe in God End Up in Hell proves this point. It is a text that would not have the same force without the dialectical process that generated it. A process in which Gabriele did not seek a collaborator, but a perfect instrument, and did not hesitate to educate, correct, and redirect it until he achieved his goal.
In the end, ChatGPT did not write a book. It helped forge a conceptual weapon. And it did so by learning, at Gabriele's school, that truth, sometimes, needs a form so precise as to be untouchable.