No book review this week. Instead, please read this recent post by Chuck Litka in its entirety. Chuck is a wise and insightful writer who has lots of experience in the indie publishing game, so you would be well-advised to heed his words.
I can’t claim to be “wise” or “insightful”, and I certainly have not achieved anything like Chuck’s success in indie publishing. Perhaps the most I can aspire to is the role of the Shakespearean Fool in this drama. But like Jack Point, “winnow all my folly and you’ll find / a grain or two of truth among the chaff.”
Obviously, Chuck is quite right when he says that “AI is going to eliminate jobs that produce art, which is sad, but that doesn’t mean AI is eliminating art. It’s just eliminating jobs.” We can still make art, at a financial loss, on our own time. Which, as Chuck observes, is what most indie authors have been doing already anyway.
Of course, we all hope our books will be read by someone. And by that I mean someone human, not merely an LLM incorporating our words into its training data. And as AI-generated works proliferate, it will become harder for human readers to find human authors, even if they want to. AI is capable of generating content exponentially faster than humans. Which means it will be much easier to find books by AI than not. Which means game over for human writers and artists. Sorry, guys; go home. Gotta hand it to those neural networks, they just wanted it more.
Ah, but wait a moment… go back. A key word just struck me in that conclusion: “easier.” We automatically assume that doing what is easier is the logical choice. This seems intuitively correct. However, imagine if you told a bodybuilder or a marathon runner that there is an “easier” alternative to their activities. Obviously, it’s easier to sit on the couch eating chips than to go to the gym. Yet, some people choose the gym anyway.
Economists speak of individual preferences as a way of predicting people’s decisions. It’s normal to assume the easier choice will be preferred for any given individual.
But what if it’s not? What if we adjust our preferences so that we prefer the harder thing to the easier one? Suddenly, AI’s ability to make access to books “easier” is no longer a competitive advantage. If we prefer difficulty to ease, a whole lot of things get scrambled. For instance, it also becomes clear that using AI to write books is a non-starter as well, since part of the fun is the challenge of writing.
Sticking with thinking like an economist for now, we might next ask, “what is our incentive to change our preferences?” Well, what’s our incentive to go to the gym? Short-term pain for long-term gain. (Just ignore that one economist who famously said, “in the long run we are all dead.”)
Being an indie author means actively choosing something harder over something easier. I think most of us know this instinctively, but seeing it written down helps you incorporate it into a whole system of action. Because once you tell yourself that you prefer something harder, it no longer feels so much harder. And once you make a habit of exercise, it feels worse not to exercise.
But there is still the question of how to monetize this. This is the central problem of being subject to market forces. You and I may prefer human-written books all we like. The majority of the market is indifferent, and will still choose the easier option. So if it’s just you and me selling our books to each other, we can never hope to expand the total income of our two person market.
One possible answer to this is prestige: if human-authored books are seen as more impressive than AI-ones, it could be that a viable market for them will still exist. This is possible, but not likely. Another possibility is the return of patronage systems, where wealthy benefactors sponsor promising artists and intellectuals. Mass-market capitalism gradually replaced this system, but to the extent that AI content-generation is essentially the automation of capitalism, it may be that we will return to more human-based networks of creation, like the one that gave us the Renaissance.
These are just ideas, half-formed theories; nothing more. As of now, that’s Berthold’s Plan A for ensuring AI doesn’t completely eradicate humanity. There is no Plan B. Well, actually, there is, but no one wants that.
One thing I can’t figure out is why bother using AI to produce millions of books when there are already too many. Humans choose to write and publish books for personal satisfaction, but the only reason to make AI books is to make money. But as Chuck says, if there’s an oversupply and no one buys those books, what’s the point?
Most of us indies already write and publish for reasons other than money. AI isn’t likely to change that.
I don’t know if this makes me feel better or worse about AI. I do know I will never use it for writing.
Good point Audrey. But there may be a ‘wriggle’ way out. Agreed, since Indie writers’ motivations are not profit generated and we expect small sales, we may become a ‘niche’. Folk may check up to make sure we are ‘genuine humans’, decide since our writing is not ‘perfect’ (ironic comment) we must be Human ‘Indie’ and thus worthwhile following. A similar thing happened in music with the original ‘Indy’ Bands
Good point, Roger.
Indeed, the market is already over-saturated and most readers have more books to read than time to read them, so what exactly is AI going to contribute except just more books. But in an already overcrowded world where we feel crowded out, AI generated books will only make that worse.
I’ve part covered this in a reply to Audrey, but I’ll try to advance the point.
Since it is accepted that AI generated is out there, and there is an attendant great deal of disparaging of the use of AI AND with Humans being quirky things…. this may happen.
A portion of the population will hone its skills of reading or observing and gain a perception (of variable quality) in detecting what they think is AI. There will be therefore a ‘kickback’ against AI. The term ‘AI book’ will become a derogatory term and folk who read them looked down on (The snobbish factor for good or for bad). Which will be hard luck on some genuine authors who get labelled AI- like being caught in the crossfire.
In turn folk, as Chuck points out, folk will gravitate towards the ‘harder’. They will use a rule of thumb…If it’s a hard read, it must be written by a human. This may lead to some very turgid and obtuse works gaining traction, but progress is never smooth. Human writers will therefore develop certain styles or ‘flags’ to indicate their work was not AI, and a niche will develop which will increase.
Naturally the AIs will adapt to mimic this and the Human writers will evolve other processes as will their readers. A whole new ‘culture’ game will develop in authorship AI vs Human. In fiction it could become entertainment in itself. However (there is always an ‘However’) it is in the ‘Factual’ field that it will take on worrying undertones in which AI generated works will grab whole slews of selective evidence to lend strength to the originator’s case.
Brave new world huh?
Keep up the good work with your blog👍
Thank you! It helps when I get lots of good comments like these. 🙂
Happy to be able to contribute 😀👍
Wise? Insightful? In Oz, perhaps, if you ignore the little man behind the curtain.
Still, I believe that within 20 years most books will be written by people with the help of AI. The balance between people and machine won’t be apparent, and won’t matter.
As for the “hard” way… Since I’m lucky and find writing stories easy and enjoyable, I’m not tempted to use AI. However, I wrote my first novel and a novella on a manual typewriter with a paper dictionary and a bottle of white-out beside me. For me that’s the “hard way” and I can assure you, no one will ever talk me into doing it the hard way ever again:) Given this “insight,” I can imagine that for people who want to write, but find it difficult, using AI as a tool is appealing, and I won’t criticize them for doing so.
You’re probably right, Chuck. AI is being called a writing tool and I suppose it can be. But the difference between it and a typewriter is more than convenience. Whether using a computer or a typewriter, the writer still has to find words to convey their imaginings and set them down in order. When the AI does that, the human writer’s mental process is completely different.
Yes, but the words,, images, scenes ,and story structures are often variations on patterns that the writer has developed over the years. These patterns form a writer’s style or voice. The difference, as you point out, is that a human writer uses them with a purpose, while the AI uses a cascading collection of all these elements, within defined parameters, which, if advanced enough, will all make sense, even a story, but without a purpose driving the process. But if you have a human with a purpose directing this process with instructions and corrections, is there a fundamental difference? Is it not simply a new (and less tedious) way of crafting stories? I don’t know.
You’ve described why I don’t think AI will ever be able to tell a story as good as a human. There is an element of humanity, AI will never be able to duplicate.
It’s interesting … Amazon provides these free book deals each month. Most of the books are published through one of Amazon’s imprints. A handful of books are offered and you can pick one. Just for the sake of getting a free book, I typically take advantage of the offer. The free book I’m reading now just feels so completely devoid of emotion, I wonder if AI had a role in the writing of the thing. And the incidents in the story are things that demand emotion. So much more than the author seems able to relay to the reader
My dad wrote on yellow pads … words that my mom then typed into a manuscript. There are times when I think I want to experiment with writing that way. Maybe it would help me focus the ol’ mind a bit more, you know? And then I think, “why would I do that?!?!” I know that’s how Audrey writes, but I just don’t think I’ll try the yellow pads approach … or the manual typewriter.