Sunny Greenhill – How to Write a Books with ChatGPT (страница 2)

You need to think not "AI helped write it," but "I created the book, using AI as a tool."

That formula is not for pretty wording. It is for the correct internal role. You are not asking AI to "make a book for you." You are using it to:

●

move through alternatives faster,

●

pull a draft out of your head,

●

expand a scene,

●

suggest intonations,

●

check logical holes,

●

find a more precise wording.

But the final form is yours.

And for that to be not just a feeling, but a practically defensible fact, you need to get used to one discipline: leaving a trace of your authorial work.

Not "folders for the sake of folders," but minimally reasonable provability: draft -> your edits -> version 1 -> version 2 -> final.

Because if tomorrow you want to explain to a publisher or partner: "I am the author," the best proof is not philosophy and not an oath… but the history of the text's creation. It shows that you did not "export a result," but created a work.

And one last, very important thing: no matter how you work with AI, responsibility for the text is always on you. Not on the model. Not on the service. On the author whose name is on the cover. And that is another indirect marker: by default, the law treats the human being as the center of the decision.

So, what have we established in this section:

You are the author not because you "ordered text," but because you created the work through decisions, editing, structure, style, selection, and responsibility. AI is a tool as long as you remain the director rather than the operator.

1.2. Where Disputes and Gray Areas Begin: "Is Your Contribution Sufficient," and Why Different Countries Look at It Differently

There is good news: in most real-life situations, the question "who is the author" is resolved calmly – the human being is considered the author, because AI cannot be an author by definition.

And there is bad news: as soon as you begin working with AI at scale, quickly, and "on a production line," a zone appears where people – publishers, lawyers, platforms, partners – start caring less that AI is not a legal subject… and start caring about something else: where exactly are you in this book?

This is where disputes appear. Not "is AI the author or not" (that is almost always closed), but whether the text contains a sufficient share of human creativity to be considered a protected human work and for your rights to it to appear stable.

I will guide you through this as a mentor, not as a lawyer: we are not going to dig through legal wording. We are going to talk about how this usually works "in the field."

When people begin to doubt authorship

Imagine that you bring a manuscript to a publisher. Or sign a contract for translation. Or sell rights to an audiobook. At some stage, a question inevitably arises: "how was the work created?"

If you say: "I wrote it," everything is simple.

If you say: "I wrote it with AI," that is not a disaster either.

But if you say: "AI wrote it, I made a few small edits," then you yourself start a chain of doubts.

Because the next thought is entirely practical: if there was little human creativity, then what exactly is being protected here as a human copyright work?

You may be certain that "it is mine anyway." But the other side cares about something else: "can we rely on this safely, legally and commercially?"

The gray word "sufficient"

Now we have reached the word that nobody likes: "sufficient."

Is your contribution sufficient for the text to be your work?

Is human participation sufficient to speak of copyright rather than "machine output"?

The problem is that "sufficient" is not an opinion and not a percentage. It is an evaluation.

And that evaluation may differ in different countries, and sometimes even among different organizations within the same country. In some places, they look at authorship more strictly: they want to see a clear creative imprint of a human being. In others, they are more pragmatic: if you organized the process and controlled the result, that is already enough.

But there is a universal logic: the more your work resembles direction and editing, the stronger your position. The more it resembles "clicked – received – published," the weaker it is.

Three levels of human participation: where are you on the scale

I am not going to overwhelm you with lists, but I will give you a clear scale.

The first level is the "operator."

You give a prompt, receive a text, and almost do not interfere. At most, you correct obvious mistakes, replace a few words, do cosmetic work.

This is the riskiest level. Not because it is "forbidden," but because in a disputed situation there is almost nothing to show: your creative contribution dissolves. The text looks like a generated product.

The second level is the "editor."

You did not just fix commas. You cut, rearrange, rewrite, align the voice, assemble the meaning, create a unified rhythm, place scenes where they belong. AI is the source of raw material, but the final form is your work.

This is already a normal professional model. Many authors work this way even without AI: they take a draft, melt it down, and turn it into a book.

The third level is the "author-director."

You have an intention, a structure, a system of characters, a dramatic framework. AI works like an accelerator: it helps with a scene variant, dialogue, description, but you are driving the entire machine of the book.

At this level, the question "who is the author" practically does not arise. Because authorship can be read from the structure and from the voice. And from the process, if it needs to be shown.

Your task is to understand honestly where you are. Not for morality. For safety and stability.

Why different countries may interpret the same process differently

Now about differences. I am not going to immerse you in legal schools, but I will explain the principle.

There are legal approaches where the central idea is that copyright protects the expression of the human personality.

In such systems, the presence of a "personal creative contribution" is especially important. If the text looks like a product of automation, they may treat it more coldly. It does not necessarily mean a ban, but in a dispute it may be more difficult.

There are more pragmatic approaches: if a person organized and controlled the creation process and made creative decisions – that is sufficient.

In reality, this means the following: if you want your book to feel stable in any jurisdiction and on any platform, you need to build the process so that it looks "human" in both the strict and the pragmatic model. In other words – do not rely on minimal participation.

The most dangerous beginner's mistake: confusing "control" with "inputting text"

A beginner often thinks: "But I wrote a long prompt, so this is my creativity."

A prompt is control, yes. But it is not always creativity that manifests in the form of expression. Especially if the prompt leads to a text that you then do not rework.

I will say it directly: a prompt by itself rarely replaces authorial work on the text. It matters as a tool, but not as proof of artistic contribution.

Creative contribution shows itself where you:

●

choose precise wording;