This post is part of a collection of thoughts I’m putting together as a companion to Five Seven Five, a stand-alone book that uses haiku I wrote as input to an AI art generator. It was a super-fun project to build. You can get it by backing my kickstarter. Or you can pre-order it at several online retailers.
Fun or not, though, the use of AI in any endeavor is filled with intrigue. So as I came closer to launching the project I decided I wanted to explore the bigger picture of how AI is impacting our world, and specifically the creative world I dwell in.
Today I’m touching on ChatGPT, Sudowrite, and the publishing environment that every writer deals with today and forever.
Before going too far, though, the image above is an interesting bit of news I came across this week. Like many, I use Word as my word processor. Apparently Microsoft added a new AI host in their last update of my machine. I have no idea what it does. But there you go. We live in an AI world.
Anyway, let’s charge on.
Yesterday I finished by noting that whether all musicians (visual artists, writers, etc.) were going to be replaced by AI or not was the wrong question. The right question, is how many would be displaced. And, let’s be clear: when I say that I mean how many of today’s artists will no longer be able to make their living because an AI is suddenly good enough.
The answer, I think, is: a lot.
That’s pretty dreary on the face of it. I think a lot of us will lose income from AI, but mostly because a lot of us cannot be—or don’t want to be—as nimble as we will need to be to make adjustments. The counter to this is that I think there will be humans who succeed as artist even after full AI implementation. This is the unfortunate nature of living in a mechanized world of capitalistic advancement. We can fight technological advancement, but…cash is king.
There is a difference for us artists, though. When <Insert Name> Company decides to install automation, all the folks doing that work lose their job. Creative people, however, live and thrive off commercial aspects of the attention economy. Which means that as long as we can gain attention from the public, and as long as the public is willing to pay for our effort, we can keep generating income. So projecting income for writers means to project how the public will seek out new work. Will they openly read AI work over their favorite human writer? Will they read both? Will they read AI for a while as a novelty, then go back to humans? Will they never really catch on to AI stories?
I don’t know where the line will fall, and neither do you. But today I’m not inclined to think readers will jump to an AI author unless that AI is both very good, and exceedingly cheap. And at present, I trust techbros to eventually make money. So, there. Make of that as you will.
Regardless, let’s make the assumption that AI does become ubiquitous and accepted by the general public, and that AI is on par with human creators. This is a likely near-term result. Writers and musicians—especially early-career independents—will be in the odd business position of seeing AI visual art being cheap, and therefore helpful, and seeing AI text creators as competition for attention. Early career independents tend to not have loads of free cash hanging around. For them, the cost side of the equation is a major impediment to their ability to gather attention. The mere existence of cheaply available AI art, then, could actually be a springboard that will help them create an artistic career. If this is true, AI will reduce the numbers of visual artists who can make a living, but increase the number of writers who can.
But … will that dynamic actually reduce the number of visual artists making a career? Written Word Media recently put out a poll of independent writers, and it was pretty clear that these early career independents rarely use outside artists today—preferring to do it all themselves. So the use of AI generated art by this class of writers will likely not impact artists to any real degree.
The world is complex, and no one actually knows anything when it comes to AI as it’s enacted today.
Focusing that on writers: No one knows, for example, if it’s true that readers don’t care about their fiction writers. If anything, I would say all the data my brain wraps around (as a mid-career mostly independent), says the opposite is true. Readers buy James Patterson. They buy (the obligatory mention of) Stephen King. They buy Nora Roberts. John Scalzi. N.K. Jemisin. Lindsay Buroker.
Occasionally they buy Ron Collins, too. But finding ways to pry attention from those writers to grab books from me is quite the challenge, eh? Having a career as a writer is about building a readership, and readers don’t just randomly select books.
This is how it has always been. This is how it will be.
So here’s another “for instance.”
The question is not what books an AI will write, but who will market them, how they will be marketed, and how hard readers will flock to them. Will that last occurrence happen in numbers large enough to kill off writers.
The answer is probably yes. It may well kill off a layer of the bottom feeders, and parts of the middle, too. People on the edge are the most vulnerable. But here’s another thing: Having seen the process work up close, I am tempted here to say something snappy like: I think the existence of traditional publishers has already killed off more writers’ careers than AI ever will.
Think about that a bit.
I recently read a fun little twitter thread that showed data that said the likelihood of a traditionally published writer still being productive five to ten years after publishing their first book was about 20%. That means 80% of the writers who were eventually published by “official” markets failed to sustain a career. Yet I know of many writers doing just fine outside the traditional markets.
Here’s another way to phrase the AI question when it comes to writers. Will the injection of AI fiction into the world affect traditionally published writers differently from independently published writers.
I don’t know, but I would think it would have to.
Traditionally published writers are already feeling the squeeze that comes with playing in corporate America. This is because they live in a hybrid world that depends not only on gaining attention, but also on that attention making enough cashflow to keep the whole corporate chain running (as well as return investment to shareholders). Every week I see new posts by mid-list authors saying they are now Officially At Wit’s End About Their Careers. If AI fiction (and AI art) drives the big publishers’ costs down, that might help them for a short while—at the expense of visual artists—but you can pretty much guarantee that will get worse in the end. Big publishing is at a real crossroads, and while it’s emotionally tough to watch writers go through that, it’s an intellectually interesting process to watch as capitalism basically eats its own tail.
Against all this comes text generators like ChatGPT and Sudowrite, the AI generators that are creating all sorts of angst among writer’s circles.
Will they put human writers out of business? Again, who get’s paid?
Mark Leslie Lefebvre recently interviewed Elizabeth Ann West, a writer fully engaged in using AI as a helper for her own creativity. Among the subtext of this conversation is that readers want books, and they want them quickly. Writers are using these AI tools not as point and click story generators, but as assistants to help focus their attention and to act as idea enhancers. How a person feels about that says a lot about the person. Assuming the writer is the final one in charge, does it matter? If it does, what about Hollywood writing, or TV writing teams, or movie scripts, or any other area where writers collaborate. Sometimes the collaborators are credited. Sometimes they are paid. Other times not. Similarly, when I have a manuscript copy edited by a human, and that copy editor suggests a hard line edit beyond a simple typo, at what point does the work stop being mine?
Financially, the answer is clear. Ethically, maybe not so much, but these are problems we have today, without AI in the conversation.
This gets even more intense when it comes to the fact that the AI we know today are trained using existing art forms. I’m planning to talk about that in more detail tomorrow, so I’m going to let that part sit for now.
My point here is that this situation with AI is no different from what we have today.
And, when we’re adding up any proposed job losses for us creatives due to attention deficits, it seems only reasonable to add in the jobs created (or held onto?) when an AI assistant helps a writer convert ideas into manuscripts that find their audience. Writers like Elizabeth Ann West say that the use of Sudowrite has essentially given them their careers. If they are the ones calling the shots, is that good?
Just as I’ve fiddled with visual art generators, I have also played around with ChatGPT. It’s an interesting little tool. I can see how it would help people, especially in the outlining stage. Today it is not particularly adroit at writing heart-grabbing fiction, but it can make a person think a bit. Which could be helpful, I suppose. At present, it’s not my thing but if you haven’t played with it, I’d suggest you do so just for the experience. I should note, too that ChatGPT and Sudowrite are two different beasts. I understand Sudowrite is more geared toward fiction, but I have not played with it to know.
For my workflow, I’d say that neither is going to directly help me. It feels clunky to my storytelling brain, so, never mind. Something that does intrigue me, though, is the idea that it could become a copy editor—focused on finding grammatical errors.
If that’s the case, then one form of copy editor could find their world squeezed.
We’ll see. I can imagine it, anyway.
Bottom line, though, if AI does become capable of writing gripping fiction—which it probably will—the fate of writers is not in the hands of coders. The fate of writers is in the hands of readers.
Which has always been true.