The Three Laws of Editing and the Coming of the AI Horde

All right. Sorry for the delay.

I’ve had a bit of a life roll recently, so this one took longer to get through than the others. In addition, the topic is tougher, so I wanted to focus better. So I took a little more time.


You might ask what the Three Laws of Editing (or the First and Only Law of Independent Publishing) have to do with the oncoming AI horde. Well. Nothing, I suppose. Or everything.

In the end, like everything else in these three connected posts, it depends on readers.

For completeness, let me start again with the Three Laws:

  1. An editor shall select those books which he or she deems will make the most money for his/her publisher.
  2. An editor shall select those books which he or she deems will provide the most entertainment value for his/her readership, in so far as doing so does not conflict with the First Law.
  3. An editor shall select those books which best serve the evolution and growth of his/her genre, in so far as doing so does not conflict with the First or Second Law.

Since the world is where it is right now, I need to state for the record that, at present — except for Five Seven Five, a single, openly made project for which the entire purpose was to explore what AI does — I do not use Artificial Intelligence for any of my work. As you might tell from this post, however, I am also not concerned with the advancing AI horde. I am not afraid of someone using my work to train an AI, for example. Nor do I care if other writers use it.

The reason I have not used AI for my work can be boiled down to two items: First, though I think copyright issues will eventually be resolved to clear the way for AI, it is worth noting that those copyright issues have not been resolved. Second, and probably more relevant, I enjoy making my own stuff up. I can imagine someday using AI as an aid, at least in the same way that, for something around thirty years, I’ve used reference books, libraries, and the Internet as a whole. Writers certainly need ideas from the outside, but the whole point of writing fiction is to be the one making the stuff up. I’m open to the idea of using anything out there to spark my thoughts, but if I felt like I wasn’t the author of something, I wouldn’t want to publish it.

Some of this reasoning is embedded deeper in the post,

I should also say I’ll be quite happy if the creatives fighting the AI horde win the game. I don’t expect they will, but I’m rooting for them and I’ve been wrong before. I’m mostly just taking what I see happening, and extrapolating it in ways I think are going to be useful going forward.

So, having waded through all this introductory stuff, let’s talk about those Laws and how the arrival might be best viewed through them.


(The Traditional Publishing Model)

The main thing that sticks in my mind about those Three Laws of Editing that Lou Anders wrote a couple of decades ago is expressly how they removed the writer from any consideration. When you think about AI, you can see exactly why those Laws have to be written that way. A Traditional Publisher truly does not care about who writes something, unless they know that writer can make money for the publisher.

This is not a complaint.

This is a fact of capitalism, and whether we like this fact or not it is the system we play in.

Until a writer (whoever they are) proves they can make enough money for the publisher, their position in the world of Traditional Publishing is tenuous. Once that writer proves they can make enough money for the publisher, their career is on more solid footing…at least until the publisher decides the writer can no longer do that. Note that this is not a quality issue. Not really. Quality is in the eye of the beholder, right? And we’ve all read traditionally published books that we think are horrible. No…the reason these Three Laws strike so deeply is exactly the fact that the First (and overriding) Law of Editing in the traditional market is simply commercial.

This means that if an AI system all by itself uses three seconds to write a book that the editor thinks will make the publisher a big chunk of money, that editor will acquire it and publish it.

This is true, and I think everyone knows it even if the idea makes them blanch, and even if every acquiring editor in the world says they wouldn’t purchase these rights. The First Law requires the editor to buy and publish books that will maximize profits. So in my mind, this is coming soon.

Of course, until an acquiring editor openly publishes a book written completely by AI, everyone is running blind. No one knows if such a book will be accepted by readers. Until hard facts exist, everything is opinion. But some traditional publisher will try, and the publisher’s readership will speak (kind of) and then we’ll know.

At least for that publisher’s readership.


(The Independent Publishing Model)

Since I first restated the Three Laws, let me start here by reiterating the One And Only Law of independent Publishing:

A writer shall create those books which they deem will provide the most entertainment value for their readership.

The reason this works as the Only Law is because the Independent model makes the writer also the acquisition editor and the publisher. Therefore, the reader provides their commercial incentive directly to the writer, and the publisher succeeds only because the writer succeeds. Duh. So you could, if you want, write a First Law that says “A Writer shall decide to create and publish those books which he or she deems will make the most money for his/her Writer.” And that’s not a bad thing, too. But it’s superfluous since in practice The One and Only Law would still directly drive it.

An Independently Published writer does not need to worry about other writers, after all. We do not need to make more money than any other writer to avoid being fired. To a much greater extent than in the world of Traditional publishing, we are in control of our careers. We don’t even have to make more money than our  past selves.

But, I should note, that we all write for different reasons.

There is, for example, a useful rule of thumb that says we writers should be aware of our primary purpose for writing, and that this primary purpose is one of three “Fs.” Finance, Fame, or Freedom. That’s not to say you can’t have interest in all three, but that in the end, the way we move along our career paths is most strongly influenced by one. Financially driven writers will focus on projects and tactics that bring them money over those that will not. Likewise, writers focused on fame and awards will try to craft their careers in those directions, and those interested in personal freedom will take a path that allows them the most leeway over their work.

How does the use of AI play into that?


Think about it.

In general, I’d say fewer independently published writers are focused on fame (which I’ll consider being focused on winning major awards or being invited to speak in front of conventions and whatnot). To be sure, there is some fame associated with the Indie world. But in the industry as a whole it’s clear we’re still mostly seen as a side-thing or a place where the less talented thrive. Still looked down the nose at, as it were. Independent writers tend not to win prestigious awards, anyway. In the end, though, this means that those few independently published writers who are driven by fame will use AI in whatever fashion will get them more attention. One can imagine several thought experiments on how that could play out, with endings that rage from heroic to tragic.

Of course, other than a few flare-ups inside the industry, the people who will directly decide if an independently published writer will become famous are the readers.

I mean…in the indie world there are only writers and readers.

No groups to carry you. No one to advocate but the writer, and no one of impact beyond the reader base themselves.

In other words, the readers will vote with their attention.

Some independent writers are financially driven. The whole idea of The Six-figure Author didn’t exist at all when there was only Trad publishing, or at least it wasn’t discussed in any way in the moderately large circle of baby writers I was in back in those days (which should maybe tell you something). Regardless, this would say that independently published authors who are financially motivated will move to AI when they can make more money with it.

Again, Duh.

Of course, what that means is that, whatever these independently published writers use AI for, it needs to help them create stories that will sell better, or (in other words) that readers who spend money want to read.

For those writers, the reader will vote directly with their wallet.

This leads us to writers who value freedom. I admit that I think this is the real hallmark of an independently published writer, probably because this is the camp that I mostly fall into. I mean, sure, I’ve wanted to make enough money (focused on Finances) to stay comfortable, and I’m not against any light shining my way (Fame). But mostly I’ve come to love being an independent writer because I like deciding what I’ll write and when I’ll write it. I like the direct contact with readers, too, of course. And, I also note that the lack of institutional deadlines from outside sources means I’ve been able to configure my life in different ways over the years without worrying about those other institutions getting in my way. In the end, I value the freedom part first.

Which, for me, is the point.

I think this is the most complex group of them all, though.

There are no homogeneous groups, of course, but I think this group of freedom-driven writers is the crew for whom the idea that “the AI world is just another tool” often fits. I can tell you that if I ever do decide to use AI as hands-on help, whatever I do with it will come filtered directly through me.

Keeping with my framework, a writer who values freedom above fame or finance will use AI in such a way as to give themselves more freedom. Reducing costs plays into this, of course. For independent writers on tight budgets, cheap AI tools could mean the difference between having a product and not having a product. But in the end, this means—first and foremost—these writers will use AI in such a way that they create more of what they want to create.

Think about what that means.

Sure, Heisenberg’s Uncertainty Principle says that it’s possible that when a writer uses an AI text generator, the output could be so good that they embrace it fully the moment it’s created (wow! That’s exactly what I would write!), but, Heisenberg aside, I can’t see it ever being that clean or perfect in real life. AI is a generator, not a mind reader. Writers have visions, and they have voices. Writers, especially those who are focused on artistic freedom, want that freedom specifically because they want to create something they see.

A writer off in the fringe will gather less attention than one with more commercial sensibilities, but as long as there’s enough attention in that fringe (a big enough readership for the writer’s frame of reference), that will work. And if that writer can use AI to get more of their ideas onto the page, that’s what they will do.

In the end, it’s still about the readers.

As it should be.

In this case, readers are voting with a combination of their attention and their pocketbooks.


The introduction of AI into creative fields is a ground shaker. No form of entertainment is immune to the changes that are, without doubt, here now. Music, visual arts, movies, and books. Everything is now exposed to the new LLM on the block.

Reactions inside the community are as varied as there are people, but I’ll split them into a couple buckets—those who are active advocates of the use of Artificial Intelligence as assistants, those who are vitriolic in their distaste for even the idea of AI, those who are moderately accepting of it (but want to get paid for any training people do with their stuff), and those who are kind of in the wings waiting for the dust to clear. As you might tell, I’m mostly in that last group.

There is, of course, another group…folks who, for the lack of a better descriptor, I’ll call scammers. These are the people who will, of course, try to make money with one-click creation of mediocre content. I’m sure some will succeed—at least in the short run. Such is the way of society in a capitalistic world. People run scams. AI will be another tool to run them with. On the other hand, if they are open about their use of AI, then are they scammers? Technically, I think not. If they are open in their use of AI, these one-click “writers” are really just acquisition editors who are working exactly like editors in Traditional Publishing houses. In other words, the First Law applies, and that editor/publisher will live and die based on (again) his or her readership.

But, of course, if they are not transparent about what they are doing, if they are pretending their one-click output is their own, then the term scam applies.

Such scams will be dealt with as such scams have always been dealt with.

That said, assuming copyright issues get resolved, and assuming AI tools exist and get better, the final question resolves into this: can AI create work of a quality that builds a readership? Can fully AI-driven entertainment succeed in the marketplace? Putting this in terms that writers think of: is an AI “better” than me (note the quotes). Or, more important, will an AI take my readership, or make it impossible to grow it?

It’s this last that is the critical question.

I could care less if people read AI things.

I care only whether they will read MY things.

Because, under either the Three Laws of Traditional Editing or The One and Only Law of Independent Publishing, the only thing that matters is what a reader (or other consumer) does. While the definition of “enough” can vary, if a writer has enough readers, that writer has a career. If not, then Houston We Have a Problem.

I’ve got plenty of competition now, so twice as much won’t matter.

As a writer, whether AI is involved or not, the challenge has always been and will always be twofold: (1) writing things that I want to write, and then (2) finding people who love to read the things I want to write.

In other words, the goal will always be to grow my readership.

The arrival of AI does not change that.

So, in every form of entertainment, the question is whether AI can actually replace a human. This puts folks in the group radically against the development and inclusion of AI into a weird little dichotomy. I understand them completely. And I don’t even argue against them. But, while they argue vehemently that AI cannot do what a human does, they are clearly worried that they will be replaced. I mean…you see the mismatch, right? If an AI truly cannot do what a human can do when it comes to the creation of art, then an AI cannot replace a human. So, unless you believe an AI actually can do just as well as a human, why worry? Putting it in terms of the Laws, if an AI cannot satisfy my readership as well as I can, why should I care?

The problem is that AI is proving to be able to replace humans for certain applications, meaning its work is already good enough for some uses. Remember, quality is in the eye of the beholder/reader.

In particular, it’s causing real issues for visual artists.

And it’s having an impact on certain aspects of the music world, as well. Video creation is almost there. Textually, AI can and already does convey basic news. Yesterday—like most days—I read several pieces of sports reporting that seemed to have been created with AI… The experience was workable. Not as fun as listening to Harry Carey used to be, but it got the job done and I was fine with it. I learned the basics of what I was looking for—scores and simple statements about who did what. If I was a true fan of any of the teams or wanted hot takes, I would have dug deeper and found a commentator I trusted.

At question, I suppose, is whether that is entertainment. How do we see the difference between journalism (just the facts ma’am), and editorial takes? And does that answer play in the world of fiction? Movies? Popular music?

Can an AI serve a purpose there?

Can it draw a readership?

No. I don’t see it.

Or, even if readers eventually do accept AI as a source of entertainment, I don’t see it impacting my world any differently than I see any other writer working today impacting it. I will still create my stuff, and it will compete in a “market” already saturated. I mean, I’m working to get my stuff seen amid ten million books out there already. What’s another ten million?

In the end, if I’m wrong and readers want AI books over mine, then that’s what they should have.

Because, in the end, that’s the thing.

The Three Laws of Editing work in the traditional world because the publisher owns access to readers. The One and Only Law of Independent Publishing works because independent writers have gained tools and processes that let us go directly to those readers instead.

If readers decide they want AI-generated content over mine, there is nothing for me to do, you know?

But I don’t think they will.

Because, in the end, I actually do believe what the doomsayers are saying. I believe that, when it comes to my readership, an AI will not be able to replace me.

Share Me
Posted in Uncategorized.

Leave a Reply

Your email address will not be published. Required fields are marked *