When the first car came out, how did the coachmen feel? When cars had been out for ten years, how did the coachmen feel? When retirement hit--voluntarily or not--how did the coachmen feel?
At this point, it's not novel or even interesting to make grand sweeping claims about AI. There have been a thousand and one essays written about the best-case scenario, the doomsday scenario, the "Goldilocks" scenario, the boring scenario... and just about every other scenario. Every other tweet about AI is either, "tHiS iS iNsAnE aNd ScArY!!" or "lol at this dumb prompt, I asked it if I should walk or drive to a car wash and it told me to fly." Everyone seems to have an agenda talking about AI--the companies push the "AI will take all jobs, so you better learn how to use it (and pay $20/month to use it, wink wink); entrepreneurs and influencers push the "AI is rapidly evolving, and if you want to stay relevant, you better learn how to use it (and pay $299 for my 30-minute course on how to use it, wink wink); and even creatives are a little guilty of "you must NEVER use AI, it will ruin your thought process and enjoyment of life (so definitely don't stay professionally relevant, stay a Luddite and fill your time with reading books/subscribing to my video)." I would love a world where AI doesn't exist, though I can't pretend I haven't used it for things like identifying what vegetables I could add more of to my diet.
Rather than making claims about what AI can or cannot do, or what AI will or will not do, I find it more interesting to discuss what would happen to my creative lifestyle under a few different scenarios. Ultimately, I am not Sam Altman, Elon Musk, or Dario Amodei; I do not even know anyone working at any of these companies, not even as a cold connection. I often sit here and say "I'll write fantasy no matter what happens." But is that true?
Or is that cope that will dissolve upon further examination?
SITUATION 1: AI Growth Declines, Stalls, or Focuses on Non-Creative Areas
The simplest case for me, I can continue to write, put my face to the world, and publish without concern. Claude's headlines about its agentic capabilities actually seem to suggest the second part of this situation is the direction of AI at the moment; Anthropic is more interested in automating mundane tasks for the office accountant or investment banker, not necessarily in developing plot analyzers (though these AIs will take a stab if you ask). Programs exist today that will claim to create a plot for you, outline your chapters, or even create a first draft. My experience with them is that, if I were to be fair, they can occasionally spark a good idea for my own outlining and writing, but they are not capable of writing better than I can. More than half the time, they just waste time I could have spent writing and outlining on my own.
In that sense, the fact that publishing as a career is so damn hard and rare is actually a sort of safety net for authors facing the doom and gloom of AI. Why try and automate a process that 99.99% of people never profit from (and can't really save money from) when there are processes where you can save tens of thousands if not hundreds of thousands of dollars from in programming, administration, and research? Again, I can't claim to understand the technical capabilities and quality of AI in these fields, but if I were the owner of OpenAI, Anthropic, Google, etc., the book industry is almost certainly not going to be my first target market.
SITUATION 2: AI Perfects the "Standard" Novel
Of all the scenarios, this one is actually the most realistic to me if we expand horizons to the next 5-10 years. AI may not necessarily target creative writing, but the downstream effects of its technological advances could make writing trope-heavy, cliche, plot-predictable stories an easy task for it. In other words, it could very well make "write to market" a thing of the past.
Whether you call it "paint by numbers," "writing to market," "meeting reader expectations," or "follow the formula," the idea of writing a book that hits all the necessary components of a certain genre is both simple to grasp and more difficult to dismiss than one might think. As much as we want to think every book is a unique work of art and unlike anything that came before it, all art--and I mean all art--draws inspiration from some art before it. The question becomes, "to what degree and how explicit?"
Recently playing Clair Obscur: Expedition 33, I noticed a ton of similarities to a game from twenty years before--Final Fantasy X. Yet when I pointed this out to a few friends, even though they ultimately agreed with me, all of them had not seen the similarities themselves. This, to me, is actually a positive; it's indicative of a game that has a clear inspiration, yet has enough creativity, passion, and personal touch such that it's a distinct work of art, not Final Fantasy X with a bunch of French names and locations.
On the other hand, I can't tell you the number of novels I've read that are blatantly derivative of other works. The most famous example might be how Sword of Shannara is essentially Lord of the Rings with different names (to be fair to Terry Brooks, his other works are much more distinct and enjoyable). I can distinctly remember reading Sword of Shannara myself, thinking "you know, the critics are kind of right," and getting to a scene mirroring the fight with the Balrog and just closing my book in frustration. Why would I read a ripoff when I had read (and watched) the original?
It's this type of work that I think is the most vulnerable to AI. We can all think of social media posts where users share an AI output of something like "write my wedding vows but in the style of Shakespeare" or "tell us the story of the Lion King but with hip hop lyrics." It's not literally the same as "write to market" but it's the same idea--take an existing structure, follow it beat-for-beat, and change up what you have only to the extent that it avoids accusations of plaigarism. This is most acute in romance, where practically every romance novel follows one of about six or so pathways to the inevitable happily ever after.
But the reality is, "write to market" is heavily suggested for a reason--it sells. So do off-the-beaten-path works, like Clair Obscur (over 5 million copies as of late February), but derivative works sell easier, like Sword of Shannara (Brooks has sold over 50 million copies of the Shannara series). Romantasy is blazing hot as a genre right now thanks to the success of Fourth Wing; I don't think you'll need to search on Amazon for long to find a hundred derivative works, all with at least a hundred reviews. From a professional writing career, there's a place for it for sure.
In my opinion, if this is the scenario that comes to pass, I believe the following would happen:
- Amazon or another entity would release a "Book For You" product wherein you type in a prompt, wait 10-15 minutes, and it has a book for you. It would be tropey, it would be cliche, it may not even be good, but it will be predictable and as you requested.
- Human authors would still have a place, but they would have to 1) balance writing to market with a little artisanal flair and 2) put themselves out in the real world to show they are real people.
- Audiences would fall into three camps--those who only read human works (which perhaps creates a rise in paperback/hardback sales, since it would be prohibitively expensive to produce and ship a "Book For You" paperback); those who read both; and those who read only "Book For You." That would fragment the authorship industry, of course, depending on how large a percent the last category is, but it wouldn't eradicate it.
Although this pathway would make it harder for me, I honestly would not hate it. The last 15 years of publishing has seen a flood of "to market" books written by ghostwriters--never mind AI--for pen names purely for profit. Epic fantasy as a genre has been less susceptible to this due to its length, complexity, and smaller market size compared to romance, but certain subgenres--litRPG, dragon-rider fantasy, romantasy--have still been hit. I will fully admit that one cannot entirely discard the concepts of "to market"; if I wrote an epic fantasy novel that didn't have world-defining stakes, magic, a big bad villain, or a traveling party of characters; if I published an epic fantasy novel with a cover that was abstract to a fault or a blurb that didn't excite the reader and hint of what to expect, it's not going to sell to my mother, let alone to actual fantasy readers. (OK, maybe my mother would buy it, but not out of love for the book itself.)
But if you told me AI forces creatives to get more creative and to stop being predictable in their plots, I wouldn't be too upset, even if it means less revenue for the write-publish-repeat ASAP crowd.
SITUATION 3: The Omnipotent AI Author
This is the doomsday scenario for creatives. This is the one where you ask an AI "write a epic fantasy series for me that merges Celtic, Norse, Greek, and Egyptian mythology into a seven-book series of over 5,000 pages, dealing with themes of duty, faith, and personal responsibility" and before the hour is up, it has something better than Game of Thrones or Harry Potter. You then follow-up with a prompt that says "great, now rewrite that book to take place in modern-day America so it mixes the feel of urban fantasy and epic fantasy" and it still pulls it off. The few who gripe about AI not having a creative soul feel hollow, shouting into a void in which no one can differentiate AI creative writing from human creative writing--nor does anyone really care.
That last sentence is actually getting at the crux of where I think the anxiety lies. When we say "will AI ever be as good as human authors?" I think what creatives are really saying is "will human readers ever care--or not care--if their books were AI produced or human produced?" I think if you could guarantee that human authors would still meet fans in the real world, sell enough to make a comfortable living, and attend DragonCon or ComicCon in 2050 with only other human attendees even while AI can simultaneously do what I described above, we'd accept that situation. We wouldn't love what AI could do, and we'd always wonder if the next generation would care about the AI/human difference, but at least in our lifetimes, we could still do what we love.
But this scenario isn't about the happy middle ground. It's about the worst case. I can spend a month writing a 200,000 word fantasy novel about dragons being hunted by men in a setting akin to Ancient Egypt, or I could spend 15 minutes prompting ChatGPT to do that. Furthermore, in this scenario, I can prompt CGPT to then create me a cover and a blurb, I can slap it up for sale, and no one--or the vast majority of readers--would care that it was made entirely by AI.
What, then?
This strips away a lot of the financial and business questions and gets down to a very simple one--why do I write in the first place?
Do I write for money? For fame? For the emotional reward that comes from writing? For spiritual purposes? For something else?
I can tell you what I would not do. I would not spend any time deliberating if a plot point or trope would impact sales, because that wouldn't matter. If I hit a hard wall in writing, to the point that I realize a series won't go anywhere (happens all the time), I wouldn't press on; I might even ask AI to finish it to satisfy my curiosity, though I could not bring myself to then charge people for it. (As a quick aside, my current, 2026 rule for AI in creative work--if you use it, you cannot charge people for it. Want AI to create a sketch of your character? I have no qualms with that, but don't then charge $14.99 for a poster of it.) I wouldn't worry about social media, because that's not something I enjoy and would gladly outsource to AI in a heartbeat if I could.
But the actual writing?
The day I realize this scenario (should it come to pass), I'll likely hit a funk. I'll probably not write for at least a couple weeks, maybe even longer, feeling nihilistic. I'll rail against AI in this blog, decry the next generation of readers, and say I was born in the wrong generation.
And then?
I'll remember how, in the late 2010s, when I had no money, maybe two dozen consistent readers, and no clear pathway to success, I wrote fantasy anyways.
I'll remember how I'd get up before sunrise--which, if you know me, is all but impossible--make my morning smoothie, let my dogs out, and write fantasy for 3-4 hours, even when sales didn't break double digits.
And I'll remember how, man, money might have been tight, economic mobility as an author seemed a pipe dream, and new releases were met with a collective shrug... but damn if it wasn't some of the most fulfilling creative time in my life.
***
Having not settled on a clear conclusion until I actually wrote this, it's funny to realize that, in some respects, I've already lived the circumstances of a "no money, no readers" lifestyle. Undoubtedly, there would be significant upheaval I'd have to make from a financial and familial perspective if Situation 3 or even Situation 2 came to pass, but those are outside the scope of this post.
I think the best way to put it is that writers write, I cannot not write, and even if the creative world is reduced to authors silently typing on keyboards in their offices, their work shared only with family and friends...
I know how the story goes.
And it's not what the market may expect.